THE DIFFERENTIAL IMPACT OF HOLOCAUST TRAUMA ACROSS THREE GENERATIONS Janine Lurie-Beck

THE DIFFERENTIAL IMPACT OF HOLOCAUST TRAUMA ACROSS THREE
GENERATIONS
Janine Lurie-Beck
Bachelor of Arts (Double Major Psychology)
University of Queensland 1994-1996
Graduate Diploma Social Science (Psychology)
Queensland University of Technology 1997
A thesis submitted for the award of Doctor of Philosophy
in the School of Psychology and Counselling
Faculty of Health
Queensland University of Technology
2007
Key words
Holocaust
survivors,
children,
grandchildren,
meta-analysis,
intergenerational
transmission of trauma, anxiety, depression, post-traumatic vulnerability, parent-child
attachment,
romantic
attachment,
family
cohesion,
separation-individuation,
communication, coping, world assumptions, post-traumatic stress, post-traumatic
growth, demographic differences.
© Janine Lurie-Beck 2007
ii
Abstract
In the current thesis, the reasons for the differential impact of Holocaust trauma on
Holocaust survivors, and the differential intergenerational transmission of this trauma to
survivors’ children and grandchildren were explored. A model specifically related to
Holocaust trauma and its transmission was developed based on trauma, family systems
and attachment theories as well as theoretical and anecdotal conjecture in the Holocaust
literature. The Model of the Differential Impact of Holocaust Trauma across Three
Generations was tested firstly by extensive meta-analyses of the literature pertaining to
the psychological health of Holocaust survivors and their descendants and secondly via
analysis of empirical study data.
The meta-analyses reported in this thesis represent the first conducted with
research pertaining to Holocaust survivors and grandchildren of Holocaust survivors.
The meta-analysis of research conducted with children of survivors is the first to include
both published and unpublished research.
Meta-analytic techniques such as meta-
regression and sub-set meta-analyses provided new information regarding the influence
of a number of unmeasured demographic variables on the psychological health of
Holocaust survivors and descendants. Based on the results of the meta-analyses it was
concluded that Holocaust survivors and their children and grandchildren suffer from a
statistically significantly higher level or greater severity of psychological symptoms
than the general population. However it was also concluded that there is statistically
significant variation in psychological health within the Holocaust survivor and
descendant populations.
Demographic variables which may explain a substantial
amount of this variation have been largely under-assessed in the literature and so an
empirical study was needed to clarify the role of demographics in determining survivor
and descendant mental health.
A total of 124 participants took part in the empirical study conducted for this
thesis with 27 Holocaust survivors, 69 children of survivors and 28 grandchildren of
survivors. A worldwide recruitment process was used to obtain these participants.
Among the demographic variables assessed in the empirical study, aspects of the
survivors’ Holocaust trauma (namely the exact nature of their Holocaust experiences,
the extent of family bereavement and their country of origin) were found to be
particularly potent predictors of not only their own psychological health but continue to
be strongly influential in determining the psychological health of their descendants.
Further highlighting the continuing influence of the Holocaust was the finding that
© Janine Lurie-Beck 2007
iii
number of Holocaust affected ancestors was the strongest demographic predictor of
grandchild of survivor psychological health.
Apart from demographic variables, the current thesis considered family
environment dimensions which have been hypothesised to play a role in the
transmission of the traumatic impact of the Holocaust from survivors to their
descendants. Within the empirical study, parent-child attachment was found to be a key
determinant in the transmission of Holocaust trauma from survivors to their children
and insecure parent-child attachment continues to reverberate through the generations.
In addition, survivors’ communication about the Holocaust and their Holocaust
experiences to their children was found to be more influential than general
communication within the family.
Ten case studies (derived from the empirical study data set) are also provided;
five Holocaust survivors, three children of survivors and two grandchildren of
survivors. These cases add further to the picture of heterogeneity of the survivor and
descendant populations in both experiences and adaptations.
It is concluded that the legacy of the Holocaust continues to leave its mark on
both its direct survivors and their descendants. Even two generations removed, the
direct and indirect effects of the Holocaust have yet to be completely nullified.
Research with Holocaust survivor families serves to highlight the differential impacts of
state-based trauma and the ways in which its effects continue to be felt for generations.
The revised and empirically tested Model of the Differential Impact of Holocaust
Trauma across Three Generations presented at the conclusion of this thesis represents a
further clarification of existing trauma theories as well as the first attempt at
determining the relative importance of both cognitive, interpersonal/interfamilial
interaction processes and demographic variables in post-trauma psychological health
and transmission of traumatic impact.
© Janine Lurie-Beck 2007
iv
Table of Contents
Key words ......................................................................................................................... ii
Abstract............................................................................................................................iii
List of Tables .................................................................................................................xiii
List of Figures................................................................................................................ xxi
Prologue ............................................................................................................................ 1
Section A .......................................................................................................................... 2
Background Literature Review of Psychological Research into the Effects of the
Holocaust on Survivors and their Descendants ................................................................ 2
Chapter One – Introduction .............................................................................................. 3
1.1. – The Holocaust as a Unique Trauma of Interest .................................................. 4
1.1.1. – The Structure of a Traumatic Event ............................................................ 5
1.1.2. – Active versus Passive Roles in Traumatic Events ....................................... 5
1.1.3. – The Rationalisation of Traumatic Events .................................................... 6
1.1.4. – The Aftermath of Traumatic Events ............................................................ 7
1.1.5. – Overview of the Assessment of the Holocaust as a Traumatic Experience 8
1.2 – The Nature of Holocaust Trauma ........................................................................ 9
1.2.1. – The Initial Phase: Gradual Removal of Civil Rights ................................... 9
1.2.2. – Phase Two: The Formation of Ghettos ...................................................... 10
1.2.3. – Phase Three: Labour and Concentration Camps ...................................... 11
1.2.4. – Alternatives to Camp Life ......................................................................... 12
1.2.5. – Immediate Aftermath: Displaced Persons Camps .................................... 12
1.3. – Thesis Aims and Rationale ............................................................................... 14
1.3.1 – Stage One: Meta-Analysis ......................................................................... 14
1.3.2. – Stage Two: Empirical Study ..................................................................... 15
1.4. – Thesis Overview ............................................................................................... 16
Chapter Two – The Study of the Impact of the Holocaust on Survivors and their
Descendants .................................................................................................................... 17
2.1. – Impact on Survivors ......................................................................................... 17
2.1.1. – History of the Assessment of Holocaust Survivor Mental Health............. 17
2.1.2. – Depression ................................................................................................. 22
2.1.3. – Anxiety ...................................................................................................... 22
2.1.4. – Posttraumatic Stress Disorder Symptoms ................................................. 23
2.1.5. – Paranoia/Fear of Further Persecution ........................................................ 23
2.1.6. – Interpersonal Trust and Intimacy............................................................... 24
2.1.7 – Factors Affecting the Severity of the Impact of Holocaust Trauma........... 24
2.1.7.1. – Coping styles and strategies. .............................................................. 24
2.1.7.2. – World assumptions. ............................................................................ 25
2.1.8. – Posttraumatic Growth ................................................................................ 26
2.2. – Impact on Children of Survivors ...................................................................... 27
2.3. – Impact on Grandchildren of Survivors ............................................................. 29
2.4. – Summary of the Impacts of the Holocaust across Three Generations.............. 30
2.5. – Critique of Research regarding Holocaust Survivors and Descendants ........... 31
2.5.1. – Definition of “Holocaust Survivor” and “Child of a Holocaust Survivor”31
2.5.2. – Nature of Control Groups .......................................................................... 32
2.5.3. – Sample Recruitment Methods ................................................................... 34
2.5.4. – Differences between Clinical and Non-clinical Study Results.................. 36
2.5.5. – Assessment of Demographic Differentials ................................................ 38
2.6. – Summary and Conclusions ............................................................................... 40
Chapter Three – The Intergenerational Transmission of Holocaust Trauma ................. 42
3.1. – Parent-Child Attachment .................................................................................. 44
3.1.1. – Survivor Parents’ Insecure Attachment to their Children ......................... 45
© Janine Lurie-Beck 2007
v
3.1.2 – Negative Effects of Insecure Attachment of Children to Parents............... 47
3.2. – Family Cohesion ............................................................................................... 50
3.2.1. – Extreme Levels of Cohesion in Survivor Families.................................... 50
3.2.2. – Extreme Family Cohesion as it Relates to the Psychological Health of
Children .................................................................................................................. 51
3.3. – Separation-Individuation .................................................................................. 52
3.3.1. – Separation-individuation problems noted among children of survivors ... 52
3.3.2. – Relationship between Separation-individuation Problems and Negative
Psychological Outcomes ........................................................................................ 54
3.4. – Communication ................................................................................................ 54
3.5. – Summary and Conclusions ............................................................................... 60
Chapter Four – Demographic and Situational Differentials in the Impact of the
Holocaust on Survivors .................................................................................................. 62
4.1. – Age during the Holocaust ................................................................................. 63
4.2. – Time Lapse since the Holocaust ....................................................................... 68
4.3. –Gender ............................................................................................................... 69
4.4. – Country of Origin ............................................................................................. 71
4.5. – Cultural Differences ......................................................................................... 73
4.6. – Reason for Survivor’s Persecution ................................................................... 73
4.7. – Nature of Holocaust Experiences ..................................................................... 75
4.8. – Loss/Survival of Family Members during the Holocaust................................. 78
4.9. – Post-war Settlement Location........................................................................... 81
4.9.1. – Europe........................................................................................................ 82
4.9.2. – Continents other than Europe. ................................................................... 84
4.9.3. – Israel. ......................................................................................................... 85
4.10. – Amount of Tme before Resettlement ............................................................. 88
4.11. – Summary and Conclusions ............................................................................. 88
Chapter Five – Demographic and Situational Differentials in the Impact of the
Holocaust on Descendants of Survivors ......................................................................... 91
5.1. – Children of Holocaust Survivor/s ..................................................................... 91
5.1.1. – Gender ....................................................................................................... 91
5.1.2. – One versus Two Survivor Parents ............................................................. 92
5.1.3. – Birth Order................................................................................................. 94
5.1.4. – Length of Time between the End of the War and the Birth of Children ... 94
5.1.5. – Birth Before or After Parental Emigration ................................................ 97
5.2. – Grandchildren of Holocaust Survivor/s ............................................................ 98
5.3. – Summary and Conclusions ............................................................................... 98
Section B ...................................................................................................................... 101
Meta-Analyses of Holocaust Survivor and Descendant Research ............................... 101
Chapter Six – Meta-Analysis Methodology ................................................................. 102
6.1. – Justification for Meta-analytic Methodology ................................................. 102
6.2. – Literature Search Methodology ...................................................................... 105
6.2.1. – Citation Sources....................................................................................... 105
6.2.2. – Process for Identifying Relevant Articles from Search Results .............. 106
6.3. – Criteria for Inclusion of Studies in Meta-analysis .......................................... 107
6.3.1. – Criteria used for Sample Selection .......................................................... 107
6.3.2. – Operationalisation of Variables ............................................................... 108
6.4. – Collection of Relevant Data/information from Individual Studies ................ 108
6.4.1. – Mean Differences between Two Groups ................................................. 109
6.4.2. – Incidence Differences between Two Groups........................................... 110
6.5. – Checks for Duplication of Results .................................................................. 110
6.6. – Calculation Methods for Meta-Analyses ........................................................ 111
© Janine Lurie-Beck 2007
vi
6.6.1. – Mean Differences between Two Groups ................................................. 112
6.6.2. – Incidence Differences between Two Groups........................................... 112
6.7. – Sub-Group Meta-Analyses ............................................................................. 112
6.8. – Meta-Regression: Correlation of Study Effect Sizes with Study Attributes .. 113
6.9. – Criteria for Inclusion of Multiple Results from Single Studies in Meta-Analysis
.................................................................................................................................. 114
6.10. – Methods for Dealing with Missing Data ...................................................... 115
6.10.1. – Missing Standard Deviations ................................................................. 115
6.10.2. – Presentation of Significance Tests without Means or Standard Deviations
.............................................................................................................................. 115
6.10.3. – Statement of Result without Data or Statistics Reported ...................... 116
6.11. – Interpretation of Meta-Analytic Findings ..................................................... 117
6.11.1. – The File Drawer Question ..................................................................... 117
6.11.2. – Testing the Homogeneity of Effect Size Sets ........................................ 118
6.12. – Overview of Meta-Analysis Section of Thesis ............................................. 119
Chapter Seven – Meta-Analyses of Survivor and Descendant Groups versus Control
Groups/General Population .......................................................................................... 120
7.1. – Method ............................................................................................................ 121
7.2. – Holocaust Survivors versus Control Groups .................................................. 121
7.2.1. – Meta-analytic Results .............................................................................. 121
7.2.2. – Studies Excluded from Meta-analyses .................................................... 123
7.3. – Children of Holocaust Survivor/s versus Control Groups .............................. 124
7.3.1. – Meta-analytic Results .............................................................................. 124
7.3.2. – Studies Excluded from Meta-analyses .................................................... 125
7.4. – Grandchildren of Holocaust survivors versus Control Groups ...................... 127
7.4.1. – Meta-analytic Results .............................................................................. 127
7.4.2. – Studies Excluded from Meta-analyses .................................................... 127
7.5. – Summary and Conclusions ............................................................................. 128
Chapter Eight – Meta-Analyses of the Moderating Influence of Demographic Variables
among Holocaust Survivors.......................................................................................... 130
8.1. – Method ............................................................................................................ 130
8.2. – Nature of Holocaust Experiences. .................................................................. 130
8.3. – Country of Origin ........................................................................................... 134
8.4. – Loss of Family Members ................................................................................ 135
8.5. – Gender ............................................................................................................ 136
8.5.1. – Meta-analytic Results .............................................................................. 136
8.5.2. – Studies Excluded from Meta-analyses .................................................... 138
8.6. – Age during the Holocaust ............................................................................... 140
8.7. – Time Lapse since the Holocaust ..................................................................... 144
8.8. – Post-war Settlement Location......................................................................... 146
8.9. – Membership of Survivor Organisations/Support Groups ............................... 149
8.10. – Summary and Conclusions ........................................................................... 149
Chapter Nine – Meta-Analyses of the Moderating Influence of Demographic Variables
among Descendants of Holocaust Survivors ................................................................ 151
9.1. – Method ............................................................................................................ 151
9.2. – Demographic Differences within the Children of Holocaust Survivor/s Group
.................................................................................................................................. 151
9.2.1. – Number of Survivor parents .................................................................... 152
9.2.2. – Gender of Survivor Parent ....................................................................... 154
9.2.3. – Type of Survivor parent’s Holocaust experiences................................... 154
9.2.4. – Parental Loss of Family Members ........................................................... 157
9.2.5. – Survivor Parent/s Country of Origin ....................................................... 158
© Janine Lurie-Beck 2007
vii
9.2.6. – Age of Survivor Parent/s during the Holocaust. ...................................... 159
9.2.7. – Length of time between the end of the war and the birth of children...... 161
9.2.8. – Location of post-war settlement. ............................................................. 163
9.2.9. – Gender ..................................................................................................... 165
9.2.10. – Birth Order............................................................................................. 167
9.2.11. – Membership of Descendants of Survivors Organisations or Support
Groups .................................................................................................................. 168
9.3. – Grandchildren of Holocaust Survivors ........................................................... 169
9.3.1. – Gender ..................................................................................................... 169
9.3.2. – Number/Gender of Child of Survivor Parent/s ........................................ 169
9.4. – Summary and Conclusions ............................................................................. 170
Chapter Ten – Refinement of the Model of the Differential Impact of Holocaust Trauma
across Three Generations based on Meta-Analyses ..................................................... 171
10.1. – Adequacy of the Assessment of Demographic Differences among Holocaust
Survivors and Descendants in the Literature ............................................................ 171
10.1.1. – Adequacy of Demographic Analysis for Holocaust Survivors ............. 171
10.1.2. – Adequacy of Demographic Analysis for Children of Holocaust Survivors
.............................................................................................................................. 172
10.1.3. – Adequacy of Demographic Analysis for Grandchildren of Holocaust
Survivors............................................................................................................... 173
10.2. – Intergenerational Differences within the Holocaust Population................... 174
10.2.1. – Direct Intergenerational Comparisons in the Literature ........................ 174
10.2.2. – Indirect Intergenerational Comparisons via Meta-analysis ................... 177
10.3. – The Need for Further Investigation .............................................................. 178
10.4. – Hypotheses for Empirical Study ................................................................... 178
10.4.1. – Hypotheses Regarding the Relationships between Model Variables .... 178
10.4.1.1. – The impact of influential psychological processes. ........................ 178
10.4.1.2. –The odes of intergenerational trauma transmission. ........................ 179
10.4.2. – Hypotheses Regarding the Influence of Demographic Variables ......... 180
10.4.3. – Hypotheses Relating to Membership of Survivor or Descendant of
Survivor Groups ................................................................................................... 182
10.5. – Summary and Conclusions ........................................................................... 183
Section C ...................................................................................................................... 185
Empirical Assessment of the Model of the Differential Impact of Holocaust Trauma
across Three Generations.............................................................................................. 185
Chapter Eleven – Empirical Study Rationale and Methodology.................................. 186
11.1. Rationale of the Empirical Study..................................................................... 186
11.2. – Method .......................................................................................................... 187
11.2.1. – Design .................................................................................................... 187
11.2.2. – Sample ................................................................................................... 187
11.2.3. – Procedure ............................................................................................... 188
11.2.4. – Translations ........................................................................................... 190
11.2.5. – Measures ................................................................................................ 191
11.2.5.1. – Depression Anxiety Stress Scales (DASS)..................................... 192
11.2.5.2. – Impact of Events Scale – Revised (IES-R)..................................... 192
11.2.5.3. – Post-Traumatic Vulnerability Scale (PTV) .................................... 193
11.2.5.4. – Adult Attachment Scale (AAS) ...................................................... 194
11.2.5.5. – Post-Traumatic Growth Inventory (PTGI) ..................................... 196
11.2.5.6. – COPE – Long Version .................................................................... 197
11.2.5.7. – Benevolence and Meaningfulness of the World sub-scales of the
World Assumptions Scale (WAS) .................................................................... 199
11.2.5.8. – Parental Care-giving Style Questionnaire (PCS) ........................... 199
© Janine Lurie-Beck 2007
viii
11.2.5.9. – Parental Fostering of Autonomy Subscale of the Parental Attachment
Questionnaire (PAQ-PFA) ............................................................................... 200
11.2.5.10. – Lichtman Holocaust Communication Questionnaire (HCQ) ....... 201
11.2.5.11. – Cohesion and Expressiveness Subscales of the Family Environment
Scale (FES) ....................................................................................................... 202
11.2.5.12. – Control questionnaire for historical influences. ........................... 203
11.2.5.13. – Omission of unresolved mourning measure. ................................ 203
11.2.5.14. – Ordering of questionnaires in questionnaire booklets. ................. 204
11.3. – Description of Sample Obtained .................................................................. 205
11.3.1. – Description of Holocaust Survivor Sample ........................................... 205
11.3.2. – Description of Children of Holocaust Survivor Sample ........................ 206
11.3.3. – Description of Grandchildren of Holocaust Survivor Sample............... 207
11.4. – Statistical Analysis Approach....................................................................... 208
Chapter Twelve – Empirical Assessment of Influential Psychological Processes and
Modes of Intergenerational Transmission Modes among Survivors and Descendants 211
12.1. – The Role of Influential Psychological Processes in Predicting Severity of
Psychological Impacts .............................................................................................. 213
12.1.1. – Influence of Coping Strategies .............................................................. 213
12.1.2. – Influence of World Assumptions of Benevolence and Meaningfulness 214
12.1.3. – Summary of the Role of Influential Psychological Processes ............... 215
12.2. – The Relationship between Posttraumatic Growth and Psychological Impact
Variables ................................................................................................................... 216
12.3. –The Role of the Proposed Modes of Trauma Transmission/Family Interaction
Variables ................................................................................................................... 217
12.3.1. – Influence of Parent-child Attachment Dimensions ............................... 217
12.3.2. – Influence of Family Cohesion ............................................................... 219
12.3.3. – Influence of Parental Encouragement of Independence ........................ 221
12.3.4. – Influence of Level of Family Communication ...................................... 222
12.3.5. – Influence of Communication about the Holocaust ................................ 223
12.3.6. – Summary of the Influence of Family Interaction Variables .................. 224
12.4. – Intergenerational Differences ....................................................................... 228
12.4.1. – Intergenerational Differences on Psychological Impact Variables ....... 228
12.4.2. – Intergenerational Differences on Influential Psychological Processes.. 230
12.3.3. – Intergenerational Differences on Perceptions of Family Interaction..... 231
12.5. – Summary and Conclusions ........................................................................... 233
Chapter Thirteen – Empirical Assessment of the Moderating Role of Holocaust
Survivor Demographic Variables ................................................................................. 239
13.1. – Demographic Variable Inter-relationships ................................................... 241
13.2. – Moderating Influence of Holocaust Survivor Demographics....................... 242
13.2.1. – Holocaust Survivor Gender ................................................................... 243
13.2.1.1 – Influence on survivor and descendant psychological health. .......... 243
13.2.1.2. – Influence on children of survivors’ perception of their parents/family
environment. ..................................................................................................... 244
13.2.2. – Holocaust Survivor Age during the Holocaust ...................................... 245
13.2.2.1. – Influence on survivor and descendant psychological health. ......... 245
13.2.2.2. – Influence on children of survivors’ perceptions of their
parents/family environment. ............................................................................. 246
13.2.3. – Nature of Holocaust Experiences .......................................................... 247
13.2.3.1. – Influence on survivor and descendant psychological health. ......... 247
13.2.3.2. – Influence on children of survivors’ perception of their parents/family
environment. ..................................................................................................... 251
13.2.4. – Loss of Family ....................................................................................... 254
© Janine Lurie-Beck 2007
ix
13.2.4.1. – Influence on survivor and descendant psychological health. ......... 254
13.2.4.2. – Influence on children of survivors’ perceptions of their
parents/family environment. ............................................................................. 256
13.2.5. – Holocaust Survivors’ Country of Origin ............................................... 258
13.2.5.1. – Influence on survivor and descendant psychological health. ......... 258
13.2.5.2. – Influence on children of survivors’ perception of their parents/family
environment. ..................................................................................................... 260
13.2.6. – Length of Time after 1945 Before Survivor Resettlement .................... 262
13.2.6.1. – Influence on survivor and descendant psychological health. ......... 262
13.2.6.2. – Influence on children of survivor’s perception of their parents/family
environment. ..................................................................................................... 263
13.2.7. – Post-war Settlement Location of Survivors ........................................... 264
13.2.7.1. – Influence on survivor and descendant psychological health. ......... 264
13.2.7.2. – Influence on survivor descendants’ perception of their parents/family
environment. ..................................................................................................... 266
13.3. – Summary and Conclusions ........................................................................... 268
Chapter Fourteen – Empirical Assessment of the Moderating Role of Descendant
Demographic Variables ................................................................................................ 272
14.1. – Demographic Variable Inter-relationships ................................................... 274
14.1.1. – Child of Survivor Demographic Variable Inter-relationships ............... 274
14.1.2. – Grandchild of Survivor Demographic Variable Inter-relationships ...... 275
14.2. – Moderating Influence of Child of Survivor Demographics.......................... 275
14.2.1. – Number of Holocaust Survivor Parents ................................................. 275
14.2.1.1. – Influence on descendant psychological health. .............................. 275
14.2.1.2. – Influence on survivor descendants’ perception of their parents/family
environment. ..................................................................................................... 276
14.2.1.3. – Perceptions of survivor versus non-survivor parents. .................... 277
14.2.2. – Time Lapse between the Holocaust and the Birth of Children of Survivors
.............................................................................................................................. 278
14.2.2.1. – Influence on descendant psychological health. .............................. 278
14.2.2.2. – Influence on descendants’ perceptions of their parents/family
environment. ..................................................................................................... 279
14.2.3. – Child of Survivor Gender ...................................................................... 281
14.2.3.1. – Influence on descendant psychological health. .............................. 281
14.2.3.2. – Influence on descendants’ perceptions of their parents/family
environment. ..................................................................................................... 282
14.2.3.3. – Interaction between parent gender and descendant gender. ........... 283
14.2.4. – Child of Survivor Birth Order ............................................................... 285
14.2.4.1. – Influence on children of survivor psychological health. ................ 285
14.2.4.2. – Influence on children of survivors’ perceptions of their
parents/family environment. ............................................................................. 286
14.3. – Moderating Influence of Grandchild of Survivor Demographics ................ 287
14.3.1. – Number of Child of Survivor Parents/Survivor Grandparents .............. 287
14.3.1.1. – Influence on descendant psychological health. .............................. 287
14.3.1.2. – Influence on descendants’ perceptions of their family environment.t
.......................................................................................................................... 288
14.3.1.3. – Perceptions of child of survivor versus non-child of survivor parents.
.......................................................................................................................... 289
14.3.2. – Grandchild of Survivor Gender ............................................................. 290
14.3.2.1. – Influence on descendant psychological health. .............................. 290
14.3.2.2. – Influence on descendants’ perceptions of their parents/family
environment. ..................................................................................................... 291
© Janine Lurie-Beck 2007
x
14.3.2.3. – Interaction between parent gender and descendant gender. ........... 291
14.3.3. – Grandchild of Survivor Birth Order ...................................................... 292
14.3.3.1. – Influence on descendant psychological health. .............................. 292
14.3.3.2. – Influence on descendants’ perceptions of their parents/family
environment. ..................................................................................................... 293
14.4. - Differences Related to Sample Characteristics ............................................. 293
14.4.1. – Membership of Survivors Organisations/Support Groups .................... 293
14.4.2. – Participation in Counselling/Therapy .................................................... 296
14.5. – Summary and Conclusions ........................................................................... 298
Chapter Fifteen – Holocaust Survivor and Descendant Case Studies .......................... 304
15.1. – Survivor Case Studies................................................................................... 304
15.1.1. – “Zosia”- Polish Child Survivor who was in Hiding .............................. 304
15.1 2. – “Siegfried”- German Child Survivor who was in Hiding ..................... 306
15.1.3. – “Greta”- Austrian Camp Survivor ......................................................... 308
15.1.4. – “Laszlo”- Hungarian Camp Survivor .................................................... 310
15.1.5. – “Hans”- German Survivor who Escaped in 1939 .................................. 311
15.1.6. – Conclusions from Survivor Case Studies .............................................. 313
15.2. – Child of Survivor Case Studies .................................................................... 315
15.2.1. – “Lena” – Daughter of two Polish Sole Survivors of the Camps ........... 315
15.2.2. – “Otto” - Son of two Dutch Survivors who were in Hiding ................... 318
15.2.3. – “Mimi” - Daughter of a Female Belgian Child Survivor who was in
Hiding ................................................................................................................... 320
15.2.4. – Summary and Conclusions from Child of Survivor Case Studies ........ 322
15.3. – Grandchild of survivor case studies ............................................................. 322
15.3.1. – “Geena” - Grandchild with One Survivor Grandparent ........................ 323
15.3.2. – “Solange” - Grandchild with Four Survivor Grandparents ................... 325
15.3.3. – Summary and Conclusions from Grandchild of Survivor Case Studies 326
15.4. – Summary and Conclusions ........................................................................... 327
Chapter Sixteen – Discussion and Conclusions ........................................................... 328
16.1. – Unique Contribution to the Holocaust Trauma Literature by the Current
Thesis ........................................................................................................................ 328
16.2. – Thought-provoking Findings Emerging from the Current Thesis ................ 329
16.2.1. – The Role of Gender ............................................................................... 329
16.2.2. – Country of Origin .................................................................................. 330
16.2.3. – The Impact of Post-war Delay in Child-rearing .................................... 332
16.2.4. – The Compounding Traumatic Impact among Survivor Dyads ............. 333
16.2.5. – The Influence of Post-war Settlement Location on post-Holocaust
Symptomatology................................................................................................... 333
16.2.6. – The Case of the Grandchildren of Holocaust Survivors ........................ 334
16.3. – Revised Model of the Differential Impact of Holocaust Trauma across Three
Generations ............................................................................................................... 335
16.4. – Applicability and Adequacy of Existing Trauma Theory in Explaining PostHolocaust Adaptation among Survivors ................................................................... 339
16.5. – Contributions to Trauma Theory by the Research Presented in the Current
Thesis ........................................................................................................................ 344
16.6. – Applicability of Attachment and Family Systems Theory in Understanding the
Intergenerational Transmission of Holocaust Trauma ............................................. 346
16.7. – Key Role played by Communication about Holocaust Experiences ............ 347
16.8. – Contemporary Needs .................................................................................... 348
16.9 – Clinical Significance and Applications ......................................................... 350
16.10. – Methodological Issues of the Current Thesis ............................................. 352
16.11. – Future Research Directions ........................................................................ 356
© Janine Lurie-Beck 2007
xi
16.12. – Conclusions ................................................................................................ 358
References .................................................................................................................... 360
Appendix A – Studies included in meta-analyses of Holocaust survivors versus
control/comparison groups with meta-regression and subset meta-analysis inclusions383
Appendix B – Studies included in meta-analyses of children of Holocaust survivors
versus control/comparison groups with meta-regression and subset meta-analysis
inclusions ...................................................................................................................... 395
Appendix C – Studies included in meta-analyses of grandchildren of Holocaust
survivors versus control/comparison groups ................................................................ 410
Appendix D – Studies included in meta-analyses of male versus female Holocaust
survivors ....................................................................................................................... 413
Appendix E – Studies included in children with one versus two Holocaust survivor
parents meta-analyses ................................................................................................... 416
Appendix F – Studies included in male versus female children of Holocaust survivors
meta-analyses ............................................................................................................... 418
Appendix G – Studies included in intergenerational comparison meta-analyses ........ 423
Appendix H – Sources of Help in Reaching Potential Study Participants ................... 425
Appendix I – Informed Consent Information Package................................................. 428
Appendix J – Multi-lingual Introduction to Study ....................................................... 431
Appendix K – Depression Anxiety Stress Scales (S. H. Lovibond & P. F. Lovibond,
1995) ............................................................................................................................. 433
Appendix L – Impact of Events Scale – Revised (D. S. Weiss & Marmar, 1997) ...... 435
Appendix M – Post-Traumatic Vulnerability Scale (Shillace, 1994)........................... 436
Appendix N – Adult Attachment Scale (Collins & Read, 1990).................................. 437
Appendix O – Post-Traumatic Growth Inventory (Tedeschi & Calhoun, 1996) ......... 438
Appendix P – COPE – Long Version (Carver et al., 1989)......................................... 439
Appendix Q – Correlations between COPE Subscales and Psychological Impact
Variables ....................................................................................................................... 441
Appendix R – Benevolence and Meaningfulness Subscales of the World Assumptions
Scale (Janoff-Bulman, 1996) ........................................................................................ 443
Appendix S – Parental Care-giving Style Questionnaire (based on Hazan and Shaver,
1986, unpublished, cited in Collins & Read, 1990) ..................................................... 444
Appendix T – Parental Fostering of Autonomy Subscale of the Parental Attachment
Questionnaire (Kenny, 1987) ....................................................................................... 445
Appendix U – Holocaust Communication Questionnaire (Lichtman, 1983) ............... 446
Appendix V – Control Questionnaire for Historical Influences ................................... 448
Appendix W – Demographic Questionnaires ............................................................... 450
Appendix X – Intercorrelations between Psychological Impact Variables within the
Empirical Study Sample ............................................................................................... 463
Appendix Y – Intercorrelations between Influential Psychological Process Variables
within the Empirical Study Sample .............................................................................. 464
Appendix Z – Intercorrelations between Family Interaction/Trauma Transmission
Variables within the Empirical Study Sample.............................................................. 467
© Janine Lurie-Beck 2007
xii
List of Tables
Table 6.1......................................................................................................................... 110
Example 2x2 contingency table for comparing incidence levels ............................................ 110
Table 7.1......................................................................................................................... 122
Summary of meta-analyses of survey/scale studies comparing survivors to control/comparison
groups ............................................................................................................................ 122
Table 7.2......................................................................................................................... 122
Summary of meta-analyses of incidence/diagnosis studies comparing survivors to
control/comparison groups ............................................................................................... 122
Table 7.3......................................................................................................................... 124
Summary of meta-analyses of survey/scale studies comparing children of survivors to
control/comparison groups ............................................................................................... 124
Table 7.4......................................................................................................................... 125
Summary of meta-analyses of incidence/diagnosis studies comparing children of survivors to
control/comparison groups ............................................................................................... 125
Table 7.5......................................................................................................................... 127
Summary of meta-analyses of survey/scale studies comparing grandchildren of survivors to
control/comparison groups ............................................................................................... 127
Table 8.1......................................................................................................................... 132
Summary of results from the literature based on the nature/type of Holocaust experiences
endured by survivors ........................................................................................................ 132
Table 8.2......................................................................................................................... 137
Meta-analysis of survey study results based on survivor gender .......................................... 137
Table 8.3......................................................................................................................... 137
Meta-regression of survivor versus control results with the female percentage of the survivor
sample............................................................................................................................ 137
Table 8.4......................................................................................................................... 137
Summary of meta-analyses of incidence/diagnosis studies comparing male and female survivors
...................................................................................................................................... 137
Table 8.5......................................................................................................................... 138
Meta-regression of incidence rates among survivor with the female percentage of the survivor
sample............................................................................................................................ 138
Table 8.6......................................................................................................................... 141
Studies assessing impact of survivor age via correlation/regression analysis ........................ 141
Table 8.7......................................................................................................................... 142
Studies assessing the impact of survivor age categorically .................................................. 142
Table 8.8......................................................................................................................... 143
Summary of meta-regression findings for average age of survivors in 1945 .......................... 143
Table 8.9......................................................................................................................... 145
Summary of meta-regression findings for time lapse since the Holocaust among survey studies
...................................................................................................................................... 145
Table 8.10. ...................................................................................................................... 145
Summary of meta-regression findings for time lapse since the Holocaust among incidence studies
...................................................................................................................................... 145
Table 8.11. ...................................................................................................................... 147
Summary of sub-set meta-analyses of survey studies by post-war settlement location among
survivors ......................................................................................................................... 147
Table 8.12. ...................................................................................................................... 147
Summary of sub-set meta-analyses of incidence/diagnosis studies by post-war settlement
location among survivors .................................................................................................. 147
Table 8.13. ...................................................................................................................... 149
© Janine Lurie-Beck 2007
xiii
Summary of sub-set meta-analyses of survey studies by sample source for survivors............ 149
Table 9.1......................................................................................................................... 152
Meta-analysis of survey study results based on number of survivor parents .......................... 152
Table 9.2......................................................................................................................... 152
Meta-regression of children of survivors versus control results with the percentage of the children
of survivor sample with two survivor parents....................................................................... 152
Table 9.3......................................................................................................................... 163
Meta-regression of children of survivor versus control results with average delay between child of
survivor birth and 1945 ..................................................................................................... 163
Table 9.4......................................................................................................................... 164
Summary of sub-set meta-analyses of survey studies by post-war settlement location for children
of survivors ..................................................................................................................... 164
Table 9.5......................................................................................................................... 164
Summary of sub-set meta-analyses of incidence/diagnosis studies by post-war settlement
location for children of survivors ........................................................................................ 164
Table 9.6......................................................................................................................... 165
Meta-analysis of survey study results based on child of survivor gender ............................... 165
Table 9.7......................................................................................................................... 166
Meta-regression of children of survivor versus control results with the female percentage of the
child of survivor sample .................................................................................................... 166
Table 9.8......................................................................................................................... 166
Summary of meta-analyses of incidence/diagnosis studies comparing male and female children
of survivors ..................................................................................................................... 166
Table 9.9......................................................................................................................... 166
Meta-regression of incidence rates among children of survivors with the female percentage of the
child of survivor sample .................................................................................................... 166
Table 9.10. ...................................................................................................................... 168
Summary of sub-set meta-analyses of survey studies by sample source for children of survivors
...................................................................................................................................... 168
Table 10.1. ...................................................................................................................... 172
Summary of the current state of evidence of the impact of survivor demographics on survivor
psychological health......................................................................................................... 172
Table 10.2. ...................................................................................................................... 173
Summary of the current state of evidence of the impact of survivor parent demographics on child
of survivor psychological health......................................................................................... 173
Table 10.3. ...................................................................................................................... 173
Summary of the current state of evidence of impact of child of survivor demographics on child of
survivor psychological health ............................................................................................ 173
Table 10.4. ...................................................................................................................... 175
Summary of meta-analyses of survey studies comparing survivors to children of survivors ..... 175
Table 10.5. ...................................................................................................................... 176
Summary of meta-analyses of survey studies comparing children to grandchildren of survivors
...................................................................................................................................... 176
Table 10.6. ...................................................................................................................... 177
Survivor, child of survivor and grandchild of survivor groups versus control groups –
Intergenerational comparison of meta-analytic effect sizes .................................................. 177
Table 11.1 ....................................................................................................................... 195
Bartholomew and Horowitz’s (1991) definitions of their four categories of adult attachment .... 195
Table 11.2. ...................................................................................................................... 198
Definitions and categorisations of COPE subscales ............................................................ 198
Table 11.3. ...................................................................................................................... 200
Reliability analysis results for the Parental Care-giving Style questionnaire ........................... 200
© Janine Lurie-Beck 2007
xiv
Table 11.4. ...................................................................................................................... 201
Reliability co-efficients for the Holocaust Communication Questionnaire and subscales ......... 201
Table 11.5. ...................................................................................................................... 204
Order and content of questionnaire booklets ...................................................................... 204
Table 12.1. ...................................................................................................................... 214
Correlations between coping strategies and psychological impact variables among Holocaust
survivors and their descendants ........................................................................................ 214
Table 12.2. ...................................................................................................................... 215
Correlations between the assumptions of world benevolence and world meaningfulness and
psychological impact variables among Holocaust survivors and their descendants ................ 215
Table 12.3. ...................................................................................................................... 216
Relationships between influential psychological processes and psychological impact variables
...................................................................................................................................... 216
Table 12.4. ...................................................................................................................... 217
Relationships between posttraumatic growth and psychological impact variables among survivors
(n = 23) ........................................................................................................................... 217
Table 12.5. ...................................................................................................................... 218
Correlations between children of survivor’s scores on psychological impact and influential
psychological process variables and their perceptions of their survivor parents (among children
with two survivor parents only [n = 51]) .............................................................................. 218
Table 12.6. ...................................................................................................................... 219
Correlations between grandchildren of survivors’ scores on psychological impact and influential
psychological process variables and their perceptions of their child of survivor parents (among
those with two child of survivor parents only [n = 10]) .......................................................... 219
Table 12.7. ...................................................................................................................... 220
Correlations between children and grandchildren of survivor/s scores on impact and influential
process variables and their perceptions of their family of origin cohesion .............................. 220
Table 12.8. ...................................................................................................................... 222
Correlations between child and grandchild of survivor scores on impact and influential process
variables and their perceptions of their survivor and child of survivor parent’s facilitation of
independence/fostering of autonomy ................................................................................. 222
Table 12.9. ...................................................................................................................... 223
Correlations between children and grandchildren of survivors’ scores on impact and influential
psychological process variables and their perceptions of their family of origin expressiveness 223
Table 12.10. .................................................................................................................... 224
Correlations between modes of communication about Holocaust experiences and children with
two survivor parents’ (n = 51) scores on psychological impact variables. .............................. 224
Table 12.11. .................................................................................................................... 226
Statistically significant relationships between proposed modes of trauma transmission/family
interaction variables and psychological impact variables and influential psychological processes
among children of survivors .............................................................................................. 226
Table 12.12. .................................................................................................................... 228
Statistically significant relationships between proposed modes of trauma transmission/family
interaction variables and psychological impact variables and influential psychological processes
among grandchildren of survivors ...................................................................................... 228
Table 12.13. .................................................................................................................... 229
Intergenerational differences in scores on psychological impact variables (including statistically
significant differences)...................................................................................................... 229
Table 12.14. .................................................................................................................... 230
Intergenerational differences in the percentage of samples scoring within normal range of tests
...................................................................................................................................... 230
Table 12.15. .................................................................................................................... 231
© Janine Lurie-Beck 2007
xv
Intergenerational differences in scores on influential psychological process variables (including
statistically significant differences) ..................................................................................... 231
Table 12.16. .................................................................................................................... 232
Mean differences in ratings of survivor parents versus child of survivor parents on parent-child
attachment dimensions and parental facilitation of independence......................................... 232
Table 12.17. .................................................................................................................... 232
Mean differences in ratings of survivor versus child of survivor families on family cohesion and
expressiveness................................................................................................................ 232
Table 13.1. ...................................................................................................................... 244
Influence of survivor gender on survivor and children of survivor scores on impact and influential
process variables ............................................................................................................. 244
Table 13.2. ...................................................................................................................... 245
Mean differences in ratings of survivor mothers versus survivor fathers among children with two
survivor parents (n = 51) on parent-child attachment dimensions and parental facilitation of
independence.................................................................................................................. 245
Table 13.3. ...................................................................................................................... 245
Children with a survivor mother versus a survivor father only perceptions of family environment
variables ......................................................................................................................... 245
Table 13.4. ...................................................................................................................... 246
Correlations between survivor age in 1945 and impact and influential process variables among
survivors and children of survivors..................................................................................... 246
Table 13.5. ...................................................................................................................... 247
Correlations between Holocaust survivor parent age and children of survivors’ ratings of survivor
parents on family interaction variables ............................................................................... 247
Table 13.6. ...................................................................................................................... 248
Holocaust survivor experience group scores on impact and influential process variables........ 248
Table 13.7. ...................................................................................................................... 250
Children of survivor scores on impact and influential process variables by survivor parent
experience groups ........................................................................................................... 250
Table 13.8. ...................................................................................................................... 251
Children of survivors’ scores on impact and influential process variables by survivor parent
experience mixture groups ............................................................................................... 251
Table 13.9. ...................................................................................................................... 252
Children of survivors’ scores on family interaction variables by Holocaust experience of survivor
parents ........................................................................................................................... 252
Table 13.10. .................................................................................................................... 253
Children of survivors’ scores on family interaction variables by survivor parent experience mixture
groups ............................................................................................................................ 253
Table 13.11. .................................................................................................................... 255
Holocaust Survivor scores on impact and influential process variables by loss of family variables
...................................................................................................................................... 255
Table 13.12. .................................................................................................................... 256
Children of survivor’ scores on impact and influential process variables by survivor parents’ loss
of family during the Holocaust ........................................................................................... 256
Table 13.13 ..................................................................................................................... 257
Children of survivors’ scores on family interaction variables by sole-survivor status of survivor
parents ........................................................................................................................... 257
Table 13.14. .................................................................................................................... 259
Holocaust survivor scores on impact and influential process variables by country of origin ..... 259
Table 13.15. .................................................................................................................... 260
Children of survivors’ scores on impact and influential process variables by survivor mother
country of origin ............................................................................................................... 260
© Janine Lurie-Beck 2007
xvi
Table 13.16. .................................................................................................................... 260
Children of survivors’ scores on impact and influential process variables by survivor father country
of origin .......................................................................................................................... 260
Table 13.17. .................................................................................................................... 261
Children of survivors’ scores on family interaction variables by survivor mother country of origin
...................................................................................................................................... 261
Table 13.18. .................................................................................................................... 261
Children of survivors’ scores on impact and influential process variables by survivor father country
of origin .......................................................................................................................... 261
Table 13.19. .................................................................................................................... 263
Correlations between survivor time in Europe before emigration and impact and influential
process variables (n = 23) ................................................................................................ 263
Table 13.20. .................................................................................................................... 263
Children of survivors born before their survivor parents’ emigration from Europe versus those
born after on impact and influential process variables ......................................................... 263
Table 13.21. .................................................................................................................... 264
Perceptions of children of survivors born before and after their survivor parents’ emigration from
Europe on family interaction variables ............................................................................... 264
Table 13.22. .................................................................................................................... 265
Holocaust Survivor post-war settlement group scores on impact and influential process variables
...................................................................................................................................... 265
Table 13.23. .................................................................................................................... 266
Children and grandchildren of survivors’ post-war settlement group scores on impact and
influential process variables .............................................................................................. 266
Table 13.24 ..................................................................................................................... 267
Children and grandchildren of survivors’ perceptions of family interaction stratified by survivor
post-war settlement location ............................................................................................. 267
Table 13.25 ..................................................................................................................... 270
Average and highest proportions of variance accounted for by survivor demographic variables
among survivor and descendant scores on psychological impact and influential psychological
process variables. ............................................................................................................ 270
Table 14.1. ...................................................................................................................... 276
Children of one versus two Holocaust survivor parents’ scores on impact and influential process
variables ......................................................................................................................... 276
Table 14.2. ...................................................................................................................... 277
Children of one versus two Holocaust survivor parents’ perceptions of family interactions ...... 277
Table 14.3. ...................................................................................................................... 278
Mean differences in ratings of survivor versus non-survivor parents on parent-child attachment
dimensions and parental facilitation of independence .......................................................... 278
Table 14.4. ...................................................................................................................... 279
Correlations between children and grandchildren of survivors’ scores on impact and influential
process variables with the time lapse between the Holocaust and the birth of children of survivors
...................................................................................................................................... 279
Table 14.5. ...................................................................................................................... 280
Correlations between children and grandchildren of survivors’ scores on family interaction
variables with the time lapse between the Holocaust and the birth of children of survivors ...... 280
Table 14.6. .................................................................................................................... 282
Female versus male children of survivors’ scores on impact and influential process variables 282
Table 14.7. ...................................................................................................................... 283
Female versus male child of survivor perceptions of their survivor parent/s on family interaction
variables ......................................................................................................................... 283
Table 14.8. ...................................................................................................................... 283
© Janine Lurie-Beck 2007
xvii
Mean differences in ratings of child-of-survivor mothers versus child-of-survivor fathers among
grandchildren with two child-of-survivor parents on parent-child attachment dimensions and
parental facilitation of autonomy ........................................................................................ 283
Table 14.9. ...................................................................................................................... 284
Statistically significantly different correlations between impact and influential process variables
and family interaction variables when stratified by child of survivor gender ............................ 284
Table 14.10. .................................................................................................................... 285
Correlations with child of survivor birth order and differences between children of survivors who
are only children and children of survivors with siblings on impact and influential process
variables ......................................................................................................................... 285
Table 14.11. .................................................................................................................... 286
Correlations with child of survivor birth order and differences between child of survivor only
children and children of survivors with siblings on family interaction variables ....................... 286
Table 14.12. .................................................................................................................... 288
Children of one versus two child of Holocaust survivor parents’ scores on impact and influential
process variables ............................................................................................................. 288
Table 14.13. .................................................................................................................... 289
Children of one versus two Child of Holocaust survivor parents scores on impact and influential
process variables ............................................................................................................. 289
Table 14.14 ..................................................................................................................... 289
Ratings of child of survivor versus non-child of survivor parents on parent-child attachment
dimensions and parental facilitation of independence .......................................................... 289
Table 14.15. .................................................................................................................... 290
Female versus male grandchildren of survivors’ scores on impact and influential process
variables ......................................................................................................................... 290
Table 14.16 ..................................................................................................................... 291
Female versus male grandchildren of survivors’ perceptions of their child of survivor parents on
family interaction variables ............................................................................................... 291
Table 14.17. .................................................................................................................... 292
Statistically significantly different correlations between impact and influential process variables
and family interaction variables when stratified by grandchild of survivor gender ................... 292
Table 14.18 ..................................................................................................................... 292
Correlations with grandchild of survivor birth order and impact and influential process variables (n
= 27)............................................................................................................................... 292
Table 14.19 ..................................................................................................................... 293
Correlations with grandchild of survivor birth order and their perceptions of their child of survivor
parents as rated on family interaction variables .................................................................. 293
Table 14.20. .................................................................................................................... 295
Comparison of survivor/descendant group member and non-members among survivors and
descendants on impact and influential process variables ..................................................... 295
Table 14.21. .................................................................................................................... 297
Comparison of survivor/descendant therapy history or no therapy history among Holocaust
survivors and descendants on impact and influential process variables ................................ 297
Table 14.22 ..................................................................................................................... 301
Average and highest proportions of variance accounted for by descendant demographic variables
among descendant scores on psychological impact and influential psychological process
variables. ........................................................................................................................ 301
Table 14.23 ..................................................................................................................... 302
Average and highest proportions of variance accounted for by descendant demographic variables
among descendant of survivors’ perceptions on family interaction variables .......................... 302
Table 15.1. ...................................................................................................................... 306
© Janine Lurie-Beck 2007
xviii
Zosia’s scores compared to whole survivor sample and relevant survivor subgroups means on
psychological impact and influential psychological process variables .................................... 306
Table 15.2. ...................................................................................................................... 308
Siegfried’s scores compared to whole survivor sample and relevant survivor subgroups means on
psychological impact and influential psychological process variables .................................... 308
Table 15.3. ...................................................................................................................... 309
Greta’s scores compared to whole survivor sample and relevant survivor subgroups means on
psychological impact and influential psychological process variables .................................... 309
Table 15.4. ...................................................................................................................... 311
Laszlo’s scores compared to whole survivor sample and relevant survivor subgroups means on
psychological impact and influential psychological process variables .................................... 311
Table 15.5. ...................................................................................................................... 313
“Hans’s” scores compared to whole survivor sample and relevant survivor subgroups means on
psychological impact and influential psychological process variables .................................... 313
Table 15.6. ...................................................................................................................... 314
Comparison of survivor case study scores on psychological impact and influential psychological
process variables ............................................................................................................. 314
Table 15.7. ...................................................................................................................... 316
Lena’s scores compared to whole child of survivor sample and relevant child of survivor
subgroups means on psychological impact and influential psychological process variables .... 316
Table 15.8. ...................................................................................................................... 317
Lena’s perceptions of her survivor father versus her perceptions of her survivor mother on gender
specific family interaction variables .................................................................................... 317
Table 15.9. ...................................................................................................................... 318
Lena’s perceptions of family interaction patterns compared to whole survivor sample and relevant
survivor subgroups means ................................................................................................ 318
Table 15.10. .................................................................................................................... 319
Otto’s scores compared to whole child of survivor sample and relevant child of survivor
subgroups means on psychological impact and influential psychological process variables .... 319
Table 15.11. .................................................................................................................... 319
Otto’s perceptions of his survivor father versus his perceptions of his survivor mother on gender
specific family interaction variables .................................................................................... 319
Table 15.12. .................................................................................................................... 320
Otto’s perceptions of family interaction patterns compared to whole child of survivor sample and
relevant child of survivor subgroups means ........................................................................ 320
Table 15.13. .................................................................................................................... 321
Mimi’s scores compared to whole child of survivor sample and relevant child of survivor
subgroups means on psychological impact and influential psychological process variables .... 321
Table 15.14. .................................................................................................................... 321
Mimi’s perceptions of her survivor mother versus her non-survivor father r on gender specific
family interaction variables ............................................................................................... 321
Table 15.15. .................................................................................................................... 322
Mimi’s perceptions of family interaction patterns compared to whole child of survivor sample and
relevant child of survivor subgroups means ........................................................................ 322
Table 15.16. .................................................................................................................... 324
Geena’s scores compared to whole grandchild of survivor sample and relevant grandchild of
survivor subgroups means on psychological impact and influential psychological process
variables ......................................................................................................................... 324
Table 15.17. .................................................................................................................... 324
Geena’s perceptions of her child of survivor father versus her perceptions of her non-child of
survivor mother................................................................................................................ 324
Table 15.18. .................................................................................................................... 324
© Janine Lurie-Beck 2007
xix
Geena’s perceptions of family interaction patterns compared to whole grandchild of survivor
sample and relevant grandchild of survivor subgroups means.............................................. 324
Table 15.19. .................................................................................................................... 325
Solange’s scores compared to whole grandchild of survivor sample and relevant grandchild of
survivor subgroups means on psychological impact and influential psychological process
variables ......................................................................................................................... 325
Table 15.20. .................................................................................................................... 326
Solange’s perceptions of her child of survivor father versus her perceptions of her child of survivor
mother ............................................................................................................................ 326
Table 15.21. .................................................................................................................... 326
Solange’s perceptions of family interaction patterns compared to the whole grandchild of survivor
sample and relevant grandchild of survivor subgroups means.............................................. 326
Table 16.1. ...................................................................................................................... 336
Status of Hypotheses relating to relationships between model variables ............................... 336
Table 16.2 ....................................................................................................................... 339
Delineation of most and least affected demographic subgroups of survivors ......................... 339
Table 16.3 ....................................................................................................................... 339
Delineation of most and least affected demographic subgroups of children of survivors ......... 339
Table 16.4 ....................................................................................................................... 339
Delineation of most and least affected demographic subgroups of grandchildren of survivors . 339
Table 16.5. ...................................................................................................................... 340
Elements from Green et al.’s (1985) and/or Wilson’s (1989) theories applied to the model of
Holocaust trauma developed in the current thesis ............................................................... 340
© Janine Lurie-Beck 2007
xx
List of Figures
Figure 2.1. Preliminary Model of the Differential Impact of Holocaust Trauma on Three
Generations ...................................................................................................................... 30
Figure 3.1. Addition of Trauma Transmission modes to the Preliminary Model of the Differential
Impact of Holocaust Trauma across Three Generations ........................................................ 61
Figure 4.1. Addition of Holocaust survivor demographic variables to the Preliminary Model of the
Differential Impact of Holocaust Trauma across Three Generations........................................ 90
Figure 5.1. Female Holocaust survivors with their babies born in a displaced persons camp.
Florence, 1946. ................................................................................................................. 95
Figure 5.2. Addition of Holocaust Survivor Descendant Demographic Moderators to Preliminary
Model of the Differential Impact of Holocaust Trauma across Three Generations................... 100
Figure 8.1. Scatterplots of effect sizes comparing survivor groups to control groups on depression
and anxiety surveys and the year studies were conducted................................................... 145
Figure 10.1. Empirical Study Hypotheses Marked on the Test Version Model of the Differential
Impact of Holocaust Trauma on Three Generations ............................................................ 184
Figure 11.1. Identification of Measures of Variables from Model of the Differential Impact of
Holocaust Trauma across Three Generations used in the empirical study ............................. 191
Figure 12.1. Section of the Test Model of the Differential Impact of Holocaust Trauma on Three
Generations to be tested in this chapter ............................................................................. 212
Figure 12.2. Scatterplots of children of survivor anxiety and depression with child of survivor
perceptions of family cohesion .......................................................................................... 221
Figure 12.3. Scatterplot of grandchildren of survivor/s Negative Attachment
Dimension/Attachment Anxiety with grandchildren of survivor/s Perceptions of Family Cohesion
...................................................................................................................................... 221
Figure 12.4. Ranking (from most important to least important) of Influential Psychological
Processes and Family Interaction Variables/Proposed Modes of Trauma Transmission in terms of
their relative importance in predicting scores on Psychological Impact Variables ................... 236
Figure 13.1. Addition of Holocaust Survivor Descendant Demographic Moderators to Testing
Model of the differential impact of Holocaust Trauma across Three Generations ................... 240
Figure 13.2. Ranking (from most important to least important) of Influential Psychological
Processes, Family Interaction Variables/Proposed Modes of Trauma Transmission and Survivor
Demographic Moderators in terms of their relative importance in predicting scores on
Psychological Impact Variables ......................................................................................... 271
Figure 14.1. Addition of Holocaust Survivor Descendant Demographic Moderators to the Test
Model of the Differential Impact of Holocaust Trauma across Three Generations................... 273
Figure 14.2. Ranking (from most important to least important) of Influential Psychological
Processes, Family Interaction Variables/Proposed Modes of Trauma Transmission and
Demographic Moderators (both survivor and descendant) in terms of their relative importance in
predicting scores on Psychological Impact Variables .......................................................... 303
Figure 16.1. Revised Model of the Differential Impact of Holocaust Trauma across Three
Generations .................................................................................................................... 338
Figure 16.2. Polish prisoners in Dachau toast their liberation from the camp (circa April/May 1945)
...................................................................................................................................... 359
© Janine Lurie-Beck 2007
xxi
The work contained in this thesis has not been previously submitted for a degree or
diploma at any other higher education institution. To the best of my knowledge and
belief, the thesis contains no material previously published or written by another person
except where due reference is made.
Signed:
Date:
© Janine Lurie-Beck 2007
______
xxii
Acknowledgements
I would like to thank my supervisors Dr Poppy Liossis and Dr Kathryn Gow for their
friendship, support, guidance and advice and my family for their love, support,
understanding and patience. I would also like to thank the survivors, children and
grandchildren of survivors who have participated in my study and other people’s
research, without whom this research would not have been possible and without whom
the world would have little conception of the true long-term impact of the Holocaust.
© Janine Lurie-Beck 2007
xxiii
This thesis is dedicated to the memory of the members of the Epstein and Lurie families
who did not survive the Holocaust and to the three special members who did and
without whom I would not exist – Jozefa Epstein and Abraham Lurie and their son Alec
Lurie. In particular, I would like to dedicate this thesis to my Nana Jo, who was proudly
waiting for me to submit this thesis, but who sadly passed away before seeing it
eventuate.
[In the camp] we used to huddle around and tell stories and I always
used to tell them about the little white house with the garden I was
going to have when the war was finished. You would probably laugh,
women in such a situation… We used to quarrel about the colour of
the walls that we were going to paint the house and what kind of
furniture we will put in and what kind of pictures will hang on the wall
and what colour roses will be in the garden. We have a marvellous
time doing all that. You could forget for a few minutes or for a couple
of hours and feel as if you were a normal human being again. Well I
wasn’t this far out. I have a little house – it is cream not white but I
have my roses in the garden all around them. So there!
Jozefa Lurie
© Janine Lurie-Beck 2007
xxiv
Prologue
I am the descendant of Holocaust survivors. My grandmother, Jozefa Lurie, was forced
from her home in the Polish city of Łodz into the city’s ghetto, passed through
Auschwitz and spent a number of years in a labour camp in Austria. My father was born
in that camp and also managed to survive.
Upon learning of their experiences in more detail, I began to think about how
well they seemed to have adapted to their experiences. I noticed differences between
my grandmother and some of her friends who I learned had also survived the Holocaust.
I also noticed differences between members of my own family.
I started to speculate as to what factors lead to better adjustment among
survivors and descendants. It was from this starting point that I began a survey of the
literature on Holocaust survivors and their families. My questions were not answered
sufficiently and so it became apparent that I would have to explore further myself to
find answers to my questions.
Janine Beck (nee Lurie)
© Janine Lurie-Beck 2007
1
Section A
Background Literature Review of Psychological Research into the Effects of the
Holocaust on Survivors and their Descendants
© Janine Lurie-Beck 2007
2
Chapter One – Introduction
It is an unfortunate aspect of the world we live in that there are a myriad of possible
traumatic events that can befall us, from natural disasters to personal assaults, wars and
terrorist attacks. McCann and Pearlman (1990, p. 10) define a traumatic experience as
being: “sudden, unexpected, or non-normative, exceeding the individual’s perceived
ability to meet its demands, and disruptive of the individual’s frame of reference and
other central psychological needs and related schemas.”
Few would disagree that the Nazi Holocaust constituted a “traumatic
experience” for its victims and survivors. Millions of men, women and children were
subjected to unspeakable horrors during this time and those fortunate to survive often
bear psychological scars which have remained with them to this day. Most people will
be at least vaguely aware of the course of events that are referred to as the Holocaust.
Images of ghettos and concentration camps are easily summonsed and have become part
of our historical psyche. While Jews were the most commonly targeted group, others
such as gypsies, homosexuals, communists, trade unionists, people with mental and/or
physical disabilities, petty criminals and many others also became victims of the Nazi
regime (Silverstrim, no date).
While the tragedy of the millions who died during the Holocaust is what lingers
in most people’s minds, it is the millions of survivors who live with the legacy of the
Holocaust that is (or should be) society’s ongoing concern. These survivors had to
piece their lives together as best they could and move on. The extent to which they
were able to do this has been the subject of much research and theoretical discussion
over the last sixty years. As Chodoff (1997) says: “We have a responsibility to
remember their ordeal, and we can still learn from their [Holocaust survivors’]
experiences as a worst-case example to guide us in dealing with the horrible examples
of mass inhumanity that we read of every day (p. 148).”
It is true that the Holocaust occurred more than half a century ago and some may
question why it is still relevant to investigate the impacts of the Holocaust. Quite apart
from the lessons that can be learnt and applied to survivors of more recent state-based
traumas is the fact that survivors and their descendants continue to this day to suffer as a
result of the Holocaust. The majority of survivors are now reaching the end years of
their life. Many are suddenly suffering from an increase of symptoms or the appearance
of symptoms from which they had, until now, been spared. The life review process that
we all face in old age is one filled with traumatic memories for survivors.
© Janine Lurie-Beck 2007
Recent
3
media coverage has pointed to the difficulties faced by survivors in aged care facilities.
These issues will be addressed in more detail in the thesis but the point needs to be
made that while the Holocaust is rapidly moving into more distant history, its effects are
still very much felt in the present and therefore worthy of continued study within the
psychological literature.
To set the context within a trauma framework, in this introductory chapter, the
Holocaust will be examined on a number of dimensions used in the assessment of
trauma. It will become obvious to the reader that the Holocaust was a unique trauma
and worthy of continued investigation. The overall aims of the current thesis are also
described in this chapter. In addition, a summary of the structure of the thesis is
provided as an overview for the reader.
1.1. – The Holocaust as a Unique Trauma of Interest
Numerous models of reaction to trauma have been developed as tools to help evaluate
the severity of a trauma and potentially predict the severity of its impact on an
individual (for example B. L. Green, 1993; B. L. Green, Wilson, & Lindy, 1985;
Wilson, 1989).
In this section, the Holocaust will be discussed in terms of a
combination of factors that were outlined in Green, Wilson and Lindy’s (1985) Working
Model for the Processing of a Traumatic Event and Wilson’s (1989) PersonEnvironment Interaction Theory of Traumatic Stress Reactions. These factors include:
structure of the trauma – single or multiple events; duration; role in trauma – active or
passive; personal exposure to intentional or non-intentional risk/threat of death or severe
physical harm; sudden violent loss of loved one/s - exposure to /witnessing death and
violence against others in particular loved ones; natural versus man-made; experienced
alone, with others or community based and; recovery environment – level of support,
societal attitudes, cultural rituals for recovery, displacement from original community.
Another key factor that was not included in either Green, Wilson and Lindy’s (1985) or
Wilson’s (1989) models is the extent to which a trauma victim or survivor can attach a
rational or palatable reason for their experience.
The application of these factors to the Holocaust will be presented in the
following subsections with key factors having lengthier discussion. It is obvious that
Holocaust survivors were personally exposed to physical harm and the threat of death to
both themselves and loved ones and that the Holocaust was a man-made rather than
natural traumatic event. These factors are shared by many other traumas. Examination
of the Holocaust on the key factors of the structure of the traumatic event, a survivor’s
© Janine Lurie-Beck 2007
4
role in the trauma, rationalisation of the trauma and recovery environment will outline
its particular uniqueness as a trauma.
1.1.1. – The Structure of a Traumatic Event
In the main, trauma tends to stem from a single event. Natural disasters such as
earthquakes, volcanic eruptions, floods, and cyclones or initiated attacks such as
terrorist attacks, rapes or muggings are over quite quickly. Such events constitute what
Solnit and Kris (1967) refer to as a “shock trauma”. They represent a “sudden peril to
life” which is momentary (Solnit & Kris, 1967). However, there are also forms of
trauma that continue for a more extended period of time. Such events include
participation in the armed forces during wartime, incarceration in a prisoner-of-war
camp or repeated episodes of abuse. Solnit and Kris (1967) introduced the term “strain
trauma” to describe this form of trauma. In other words, the form of trauma continues
for an extended period and places the individual under strain.
Survivors of the
Holocaust can certainly fall into this latter category of “strain trauma”.
Strain trauma is viewed as potentially more damaging to the psyche than shock
trauma. Bistritz (1988) argues that even the most harrowing of experiences can be
handled if what is occurring is “predictable and potentially limited in time.” In other
words, if the person can see an end to their suffering they have a heightened capacity to
deal with the trauma as it is occurring because it has a foreseeable end. The nature of
the events of the Holocaust meant that there was a high degree of uncertainty and
unpredictability of what would happen on a daily basis as well as the overall chances of
survival. Bistritz (1988) argues that given this background, survivors were unable to
psychologically defend themselves as much as survivors of forms of shock trauma are
able.
1.1.2. – Active versus Passive Roles in Traumatic Events
There is also the issue of active versus passive roles within traumatic experiences.
Military personnel can be seen to have played an active role in the events leading up to
and within their trauma events. Holocaust survivors on the whole (excepting those
persecuted for resistance) played a passive role. They had not actively participated in
circumstances leading to their ordeals; they were simply forced to obey their
captors/persecutors.
This
passivity
would
have
also
contributed
to
the
incomprehensibility of their experiences. For example, Sigal and Adler (1976) found
Canadian veterans who had been in Japanese prisoner-of-war camps showed better postwar adjustment than Jewish concentration camp survivors. He argued that this was
© Janine Lurie-Beck 2007
5
partially because the soldiers had played an active fighting role compared to the Jewish
camp survivors who were not interned for anything active.
As Tas (1951/1995) pointed out, soldiers in action were able to relieve pent up
tensions and aggression via combat, an outlet that a Holocaust survivor was missing.
Tas (1951/1995) suggested that because any expression or venting of emotion was a
highly dangerous action for a survivor (because of reprisals for the individual and often
for many surrounding them), their suffering was compounded by the enforced
accumulation of feelings such as anxiety and rage. The mechanism of repressing these
feelings during the Holocaust as a survival strategy would then become a hindrance
when the cathartic release of these emotions was almost impossible for many survivors
in the post-war period (Tas, 1951/1995).
Recent research in Australia has found that playing an active role can minimise
traumatic impact. Parslow (2005) stated that people caught by the Canberra bushfires
who were allowed to play a role in trying to protect their homes and were given more
warning of the need to evacuate suffered fewer post-traumatic symptoms than those
forced to evacuate with minimal notice. Weisaeth (2005) reported that survivors of a
number of industrial accidents in Scandinavia suffered from fewer symptoms if they
had played an active role in trying to rescue fellow workers.
1.1.3. – The Rationalisation of Traumatic Events
It can be argued that while members of the armed forces who participated in combat
and/or were interned in prisoner of war camps certainly endured hardships for years,
they were able to assign a more palatable reason to their suffering than Holocaust
survivors were to their experiences.
Military personnel have a clear mission and
ultimate goal which leads them to believe that their actions and their suffering are
justified. They are involved in fighting an “enemy” which perhaps threatens their
country or their way of life in some way. There is a clear cause and effect between their
actions and their experiences. Even for prisoners of war who are not in control of the
events that occur to them as much as they were when in combat can at least understand
the reason for their internment. They can rationalise that they have been captured and
detained because they are enemies of their captors and were fighting against them.
Holocaust survivors in the main were subject to persecutions and incarceration
in camps because of their nationality or ethnicity. They had committed no act that
could be considered an attack against their captors.
For Jews and Gypsies their
internment was aimed at the complete destruction of their race based on the opinion that
they were a lesser group in society and not deserving of equal rights or even life. This
© Janine Lurie-Beck 2007
6
must have been a very difficult concept to grasp and survivors must have felt a greater
indignation and incomprehension at their detention than prisoners of war.
Support for the notion that a more readily derived rationalisation of a traumatic
event can be protective was discussed by Makhashvili, Beberashvili, Tsiskarishvili,
Kiladze, and Zazashvili (2005). Makhasvili et al. (2005) reported that PTSD rates were
much higher among Georgian civilians displaced during the 1990s due to internal
conflict in Georgia (61%) compared to ex-Chechnyan fighters (11%) and ex-political
prisoners (14%) who were persecuted for support of Georgian ex-president
Gamsakhurdia who was ousted in 1992.
In another study from the Baltic states,
Kazlauskas, Gailiene, & Domanskaite-Gota (2005) reported higher levels of anxiety and
depression among deported civilians than ex-political prisoners in Lithuania.
The
political prisoners could more readily explain their persecution (active opposition to
their captors) whereas the displaced and deported civilians could not.
1.1.4. – The Aftermath of Traumatic Events
How a survivor is treated and received not only by their immediate friends and families
but by society as a whole has been cited by a number of researchers as a key factor in
determining their post-trauma adjustment (Bower, 1994; de Silva, 1999; Gill, 1994; B.
L. Green, 1993; B. L. Green et al., 1985; Kestenberg & Kestenberg, 1990a, 1990b;
McCann & Pearlman, 1990). Symonds (1980) talks about the notion of “second injury”
that occurs to survivors when they are greeted by unsupportive or blaming reactions
from others. It is suggested that such a reception can be almost as damaging as the
trauma itself. Certainly, it is a widely accepted view that the less than supportive
reception Vietnam veterans received upon their return home did nothing for their
recovery (de Silva, 1999; Lomranz, 1995; McCann & Pearlman, 1990; Z. Solomon,
1995). Sigal and Adler’s (1976) Canadian prisoners-of-war were also theorised to have
adjusted to post-war life because they returned to a homeland that honoured their ordeal.
Reactions to Holocaust survivors were largely dismissive. Brom, Durst and
Aghassy (2002) talk of the horror people experienced when they learned of what
survivors had endured. People did not want to think about what had happened and so
largely shut survivors out rather than reaching out in an effort to support them and help
them recover. Kestenberg and Kestenberg (1990b) note that soon after large scale
natural disasters organised attempts to help people and support them through the initial
aftermath are put into place. Counselling services are often part of this process. While
it cannot be denied that survivors were provided with a degree of material assistance
upon their liberation, an organised attempt to provide psychological help was sadly
© Janine Lurie-Beck 2007
7
lacking (Friedman, 1948). A number of researchers have suggested that the availability
of social supports and networks in the survivors’ community are a key factor in
determining their long term recovery (de Silva, 1999; B. L. Green, 1993; B. L. Green et
al., 1985).
While Vietnam veterans may not have had societal support, they were able to
return to family and friends. For a large number of Holocaust survivors there were no
surviving relatives to seek comfort from, no friends, no home and a community
unwillingly to welcome them back with open arms (Eitinger, 1973; Gill, 1994). For a
majority of Holocaust survivors, there was very little to come home to (Rappaport,
1968). Many were the sole survivors of their families, their homes and property had
long been confiscated and in most cases was irretrievable.
Returning to such a
“vacuum” compounded the trauma for these survivors (Kestenberg & Kestenberg,
1990b; Zolno & Basch, 2000). As Davidson (1980a) states: “The destruction of the
community and the previous life of the survivor damaged his basic sense of security and
undermined his identity and sense of continuity with the past.”
Survivors whose
communities were dismantled or destroyed must have been affected by the resulting
isolation.
Indeed, the intactness of the survivor’s community is included in Green et
al.’s (1985) Working Model for the Processing of a Catastrophic Event as an important
aspect of the recovery environment. A person’s sense of connection to their community
has been statistically significantly linked with psychological well-being (for example
W. B. Davidson & Cotter, 1991).
The loss of community was felt so keenly by some survivors that even in the
displaced persons camps, that survivors found themselves in immediately after the war,
they began to form social groups with survivors of the same region or town as their
own.
Such groups appeared in much larger numbers when survivors resettled in
countries such as America. Participation in such groups provided at least a small form
of continuity with their destroyed pre-war lives (Zolno & Basch, 2000).
1.1.5. – Overview of the Assessment of the Holocaust as a Traumatic Experience
While there are certainly some traumatic experiences that share some commonalities
with the Holocaust, it is unique in its combination of traumatic elements.
The
Holocaust can be defined as a strain trauma of long duration, for which there was no
readily acceptable rationalisation or reason, which was followed by little societal
support for victims and in many cases indifference and which dramatically altered a
survivors’ life course. Not only was the experience itself horrific, but the survivor had
nothing to return to and was forcibly severed from the life course they were on. Similar
© Janine Lurie-Beck 2007
8
events have unfortunately occurred not only in Europe (Former Yugoslavia), but in Asia
(Cambodia) and Africa too (Rwanda and Sudan) (Chodoff, 1997). Attempted genocides
are, sadly, not solely the domain of history. Holocaust survivors provide us with the
unique opportunity to examine the long term impacts of attempted genocides as well as
intergenerational effects. As Wieland-Burston (2005) puts it:
Although the Shoah was a unique experience in itself, it nevertheless
serves as a prototype for genocide, for massive collective trauma and
its long-term effects on a population. In this respect, research on the
topic is of vital importance for all who work in the field of
psychotherapy, not only for us today, living and dealing with the
direct descendants of the Shoah, but also for the future. We cannot
suppose that genocide and mass collective trauma are a thing of the
past (p. 513).
1.2 – The Nature of Holocaust Trauma
By the end of World War II, roughly eleven million people had died at the hands of
Hitler’s Third Reich (Simon Wiesenthal Centre, n.d.)
However, the Nazis were not
successful in completely wiping out the societal groups that they had sought to. By the
end of 1945 there were one and a half to two million displaced persons who had
suffered to some degree at the hands of the Third Reich but had managed to survive.
There are a myriad of texts available for the reader in search of detailed accounts
of what the Jews and other targeted groups went through during World War II. The
current section merely gives a broad overview of the types of experiences these people
had to endure to serve as a context for the understanding of their post-war adjustment.
1.2.1. – The Initial Phase: Gradual Removal of Civil Rights
From the moment Hitler was declared German Chancellor on 30 January 1933,
the gradual removal of rights for German Jews began (Edelheit & Edelheit, 1994).
Initially these were only aimed at causing humiliation but eventually they impacted on
the ability to survive. The long list of humiliations included the boycott of Jewishowned businesses, the barring of Jews from government service such as education, the
restriction, and later complete barring, of Jewish children from attending German
schools and universities, playing in playgrounds (1933), followed by removal of
citizenship (1935), the banning of Jews from public streets on certain days, forbidding
Jews to have driver’s licenses or car registration, forcing Jews to sell their businesses,
property, investments and jewellery to the Reich at artificially low prices, forcing them
to carry an identification card (1938), forcing Jews to relinquish radios, cameras and
other electric objects to the police, providing them with restrictive food rations denying
them things such as meat and milk and limiting amount of clothing (1939) (Holocaust
© Janine Lurie-Beck 2007
9
time line 1933-1945, n.d.).
Edelheit and Edelheit (1994) note that by the beginning of
the war at least 121 laws were passed, each of which denied German Jews a
fundamental right. These restrictions of freedom were extended to the Jews of Austria
after the Anschluss in 1938 and Czechoslovakia in early 1939 (Edelheit & Edelheit,
1994).
As the German Army marched forward into Poland in September 1939, sparking
the beginning of World War II, the Jews of all occupied territories soon joined their
German, Austrian and Czech counterparts in becoming bureaucratically marginalised
members of society. Additional indignities were added such as the freezing of Jewish
back accounts, the introduction of a curfew and compulsory Star of David armbands
(Holocaust time line 1933-1945, n.d.).
Groups that later suffered at the hands of the Third Reich (such as political
prisoners) did not endure these earlier years of incremental rights removals. This
gradual removal of civil rights and erosion of dignity was largely unique to the Jews
(although it was also applicable to Gypsies to a lesser extent) and also lasted for
differing lengths of time depending on the country.
German Jews had begun to
experience such changes from six years prior to the onset of war while Jews of other
countries retained such rights until their countries were invaded during the war.
Germany invaded Belgium, Denmark, France, Luxembourg, the Netherlands and
Norway in 1940, Yugoslavia and Greece in 1941 and Hungary only in 1944 after
initially being a German ally (Holocaust time line 1933-1945, n.d.).
1.2.2. – Phase Two: The Formation of Ghettos
In Nazi-occupied Poland, Jewish ghettos were established from 1939 onwards
(Holocaust time line 1933-1945, n.d.). In large cities, such as Warsaw, Lodz and
Krakow, sections of the city were closed off to gentile residents and a walled ghetto was
created to house the cities’ Jewish populations. Often, Jews from surrounding areas
were also moved into the ghettos and, eventually, Jews from other countries occupied
by the Nazis were transported to these ghettos as well.
According to Berenbaum (1993), ghetto life was one of “squalor, hunger,
disease and despair”. Accommodation was very cramped with ten to fifteen people
occupying a space previously used by no more than four. When moving into the ghetto,
people were forced to leave most of their belongings behind and could take only
whatever they could pile onto a wagon. Ever diminishing food rations saw the ghetto
populations gradually starving (Rosenbloom, 1988).
Serious public health problems
led to epidemics of diseases such as typhus (Berenbaum, 1993).
© Janine Lurie-Beck 2007
10
Despite the dramatic drop in living standards, inhabitants of the ghettos got on
with life as best they could. Adjustments were made to a life of daily struggle for food,
warmth, sanitation, shelter and clothing. Every effort was made to continue cultural life
with clandestine schools created and religious services and entertainment such as music
and theatre productions arranged to create at least a semblance of normal life
(Berenbaum, 1993).
However, ghetto inhabitants lived in constant fear as residents were soon drafted
for ‘labour conscription’ and ‘deportation’. In the European summer of 1942, the
ghettos of Eastern Europe began to be cleared of inhabitants, which in the majority of
cases involved the separation of families. Two years later more than two million Jews
had been transported to concentration camps and there were no ghettos left in Eastern
Europe (Berenbaum, 1993).
1.2.3. – Phase Three: Labour and Concentration Camps
Life in a concentration camp can be described as nothing less than horrific. People only
avoided the gas chambers by being fit to work.
Bluhm (1948) explains how people
interned in camps were forced to work sixteen hour days. The work was often backbreaking and anyone who did not fulfil their quota or was considered to be not pulling
their weight was severely punished by beatings, torture or frequently death (Bistritz,
1988; Bluhm, 1948). Inmates were provided with very little food and often inadequate
clothing with which to sustain themselves in order to do this work (Bluhm, 1948).
People died from starvation or diseases resulting from the unsanitary living conditions
as well as at the hands of the Nazis (Bistritz, 1988). Inmates dealt with death on a daily
basis and often witnessed the deaths of family members, friends and fellow inmates,
often in brutal circumstances (Bistritz, 1988). A number of camp inmates retreated into
a state of “psychological hibernation” because of the extreme shock of camp
incarceration and all that this entailed. They became apathetic and lost the will for selfpreservation and in most cases did not survive for long. Camp inmates who fell victim
to this apathy were often termed “musselmen” (Chodoff, 1997).
Despite all this many were able to survive for many years (Bluhm, 1948),
although a great number tragically died after liberation as a result of the conditions they
endured. They were able to survive largely because they were the most fit and able in
the first place. The very young and the very old as well as anyone with an obvious
illness or physical disability were immediately sent to the gas chamber on arrival.
Bluhm (1948) also notes that conditions differed from camp to camp with the death
rates obviously higher in camps with the worst conditions.
© Janine Lurie-Beck 2007
11
1.2.4. – Alternatives to Camp Life
Some of the Jews, and other groups targeted by the Nazis, managed to escape from
ghettos and/or camps and managed to survive in hiding or with an assumed identity.
Some had the opportunity of joining partisan groups and therefore join in the fight
against the Nazis.
Rosenbloom (1988) explains how some were provided with a hiding place such
as a secret compartment in an attic or basement in the home of a sympathetic Christian.
Those lucky enough to secure such a hiding place were sometimes able to survive for
quite extensive periods, even years, which is no mean feat when considering the often
cramped confines of their sanctuary.
Other Jews survived by assuming an “Aryan” identity with the help of forged
papers. The success with which this was done was to some extent dependent on the
person’s ability to fit in with their assumed persona with a convincing accent and roleappropriate behaviour. Of course, Rosenbloom (1988) points out that there was also a
degree of luck in how long an assumed identity could be maintained. There were many
obstacles to overcome, not the least of which being the willingness of some to inform
authorities of likely impostors for monetary reward (Rosenbloom, 1988).
However, both groups shared the constant fear of discovery which, in
Niederland’s (1968) experience, often led to great levels of post-war anxiety.
Rosenbloom (1988) suggests that “the constant necessity to remain alert and watchful
produced a unique set of psychological stresses”.
The current author would suggest that survivors who joined partisan groups may
well have had a psychological edge as they would have had the satisfaction of fighting
back against the Nazis.
While these survivors were also under threat of being
discovered, they were in groups and could therefore look out for each other thereby
dissipating some of the fear of being found. Survivors in hiding were often on their
own with no one to help them feel more secure or protected. Despite these hardships
survival in hiding was much more likely than survival of the camps. Vogel (1994)
quotes data that suggests that between 400,000 and 500,000 Jews survived in hiding,
while no more than 75,000 survived the camps.
1.2.5. – Immediate Aftermath: Displaced Persons Camps
The suffering was not over for all Holocaust survivors once hostilities ceased in 1945.
While survivors from Western European countries such as France, Belgium and
Holland were, on the whole, willing to return home and able to reintegrate into the
© Janine Lurie-Beck 2007
12
community with little difficulty, survivors from Eastern Europe were far less willing to
return, and with good reason.
A large number of survivors were met with animosity rather than sympathy and
understanding in their homelands (L. Berger, 1988; Bergmann & Jucovy, 1990; Danieli,
1988; Edelheit & Edelheit, 1994; Johnson, 1995). Bulgarian, Hungarian, Polish and
Romanian Jewish survivors were greeted with widespread anti-semitism which
frequently led to bloodshed. For example, in post-war Poland, during the first seven
months after the war, 350 Jews died as a result of anti-semitic violence (Gilbert, 1987;
Johnson, 1995). Survivors often found it near impossible to reclaim their possessions
and homes from the people who had occupied them after they had been forced to leave
by the Nazis (Kestenberg & Kestenberg, 1990b).
It is therefore understandable that a large proportion of these survivors wanted to
get far away from Europe, to America or to the soon to be established state of Israel. A
massive refugee problem soon developed with hundreds of thousands of displaced
persons wanting to emigrate to countries reluctant to take them in (at least not in large
numbers) (L. Berger, 1988; Gill, 1994; Kestenberg & Kestenberg, 1990b).
The
situation for those wanting to emigrate to Palestine/Israel was particularly difficult with
the British Administration very reluctant to allow sizeable refugee shipments to enter
Palestine (until the establishment of the state of Israel on 15 May, 1948). It came to the
stage where the British navy either sent refugee boats back to Germany or Italy or rerouted them to Cyprus where large numbers of displaced persons (of whom a very large
percentage were Jewish Holocaust survivors) were housed in what was, for all intents
and purposes, another concentration camp: a detention centre opened in August 1946
(L. Berger, 1988; Bergmann & Jucovy, 1990; Edelheit & Edelheit, 1994; Friedman,
1948, 1949; Kestenberg & Kestenberg, 1990b). This was obviously a highly insensitive
method of operation and a potentially further traumatising experience for the survivors
interned there. On his numerous visits to the displaced persons (DP) camps in Cyprus,
Friedman (1948; 1949) noted that “the barbed wire of Cyprus cruelly result[ed] in
bringing back to many of the immigrants, either by association or by the re-stimulation
of ingrained, conditioned responses, some of the behaviour patterns of the
Concentration Camps. Anxieties which were held in abeyance for many years now
erupt[ed] to the surface.” He suggested that:
To attempt to revive the emotional life of people who continue to live
in the precarious and threatening atmosphere of the DP camps in
Europe or incarcerated on the purgatorial island of Cyprus, hemmed in
by barbed wire and armed guards in an ominous re-enactment of the
© Janine Lurie-Beck 2007
13
German chapter of their torture, might appear an impossible
undertaking. (p. 507)
It took over ten years for some, but finally, the remaining survivors were able to
leave Europe and set up new lives far away from the site of their suffering. The
situation was eased by the establishment of the state of Israel in 1948 and legislation
enacted in the United States Congress shortly afterwards admitting a large number of
displaced persons to America (L. Berger, 1988). There are numerous and varying
estimates of the number of emigrants to the various countries around the world (Krell,
1997a; United States Holocaust Memorial Museum, n.d.). Hass (1996) suggests that
55% of Jewish survivors emigrated to Israel, 25% emigrated to the United States, 10%
emigrated to Canada, Australia, South Africa or Argentina and the remaining 10%
stayed in Europe. The last DP camp closed on 28 February 1957 (Gill, 1994).
1.3. – Thesis Aims and Rationale
The overarching aim of the current thesis is to identify which sub-groups of Holocaust
survivors and their descendants display the most psychological and inter-personal
problems. Who are the most resilient and who are the most vulnerable? To this end,
factors that differentiate between well-adjusted and less-adjusted individuals are of
interest. These factors include many which are demographic and/or situational nature.
To identify these factors, two approaches are utilised. A meta-analytic review of the
literature is undertaken with the aim of devising a model of the differential impact of the
Holocaust. The development of this model is informed by a number of theoretical
frameworks including trauma theory, attachment theory and family systems theory. An
international, transgenerational study is then conducted to test this model and the model
is then revised based on the data obtained from this study.
1.3.1 – Stage One: Meta-Analysis
In the first approach, a detailed review and meta-analytic collation of the psychological
and psychiatric research conducted with survivors and their descendants (up to and
including the year 2006) is undertaken.
Within this stage, the extent of existing
empirical support for the factors theorised to differentiate between well- and lessadjusted survivors and descendants are established. The results of these meta-analyses
are used to revise a preliminary model outlining the impact of the Holocaust on
survivors and their descendants and variables that influence the severity of that impact.
Meta-analysis is also used to determine the size and range of differences
between survivor and descendant samples and control/general population samples. It is
hypothesised that, overall, survivor groups will score higher than the general population
© Janine Lurie-Beck 2007
14
on pathological variables, and evidence higher levels of maladaptive familial and
interactional patterns.
The aim is to demonstrate that there is a wide range of
differences between survivor groups and the general population (via the measurement of
control groups) which reflects a wide range within the survivor and descendant
populations themselves and that they do not represent a uniform, homogeneous group.
Therefore, research should focus on differences within the survivor population and what
factors explain those differences, rather than finding evidence for the obvious point that
survivors and their descendants differ from the general population (which in fact has
been established over the past six decades).
There is little argument that survivors and descendants differ in terms of their
psychological health and the way in which they relate to others. However, there has
been little focus on identifying the factors that explain these differences and which
factors are the most determinant. As a result of the meta-analyses and literature review
conducted, a draft model of the differential impact of the Holocaust will be constructed
which delineates numerous demographic and situational variables that appear to
moderate the impact of the Holocaust on survivors and descendants.
1.3.2. – Stage Two: Empirical Study
The empirical study conducted for the second phase or stage of the research follows on
from the meta-analyses. It aims to test the veracity of the model developed from the
meta-analyses and literature review.
In addition, many variables that have been
theorised to explain differences in post-war adjustment have been inadequately assessed
to date and some have not been assessed at all. Therefore a secondary aim of the study
is to provide some data on these variables in an attempt to clarify their impact. Because
many of the variables included in the meta-analysis were assessed in isolation in the
existing literature, the study conducted for the current thesis represents a unique
opportunity to determine the relative importance of these potential demographic and
situational moderators.
Analysis of the data obtained in this empirical study provides the reader with an
idea of the veracity of the model of the differential impact of the Holocaust. In addition,
it delineates the demographic variables that appear to be the most influential in the
severity of symptoms experienced by survivors and their descendants.
To further
illustrate the heterogeneity of the survivor and descendant populations, ten case studies
are also presented.
© Janine Lurie-Beck 2007
15
1.4. – Thesis Overview
A summary of the main psychological impacts of the Holocaust on survivors
and their descendants is presented in Section A. Given that the traumatic impact of the
Holocaust is seen to have been transmitted through the generations, the main arguments
as to how Holocaust trauma can be transmitted intergenerationally are also summarised
(Chapter Three). Chapters are also included that summarise current thoughts on the
impact of numerous demographic and situational variables on the three generations as
well as on the transmission processes (Chapters Four and Five).
Section B starts with an explanation of, and justification for, the meta-analytic
methodology used to review the current state of empirical data in the Holocaust
literature (Chapter Six). The first meta-analytic results chapter examines data that
compare Holocaust survivors or their descendants to control/comparison groups in the
literature (Chapter Seven). The next two chapters present meta-analytic and non-metaanalytic reviews of the current state of the literature with regards to demographic subgroups of the Holocaust survivor and descendant populations (Chapters Eight and
Nine). Finally, the last chapter in this section revisits the draft model presented in
Section A and revises it based on the meta-analyses presented in the chapters in Section
B (Chapter Ten).
Section C reports on the empirical study designed to test the revised model from
Section B. Chapter Eleven summarises the rationale and methodology of the study.
Chapter Twelve reports on inter-relationships between model variables, while the next
two chapters (Chapters Thirteen and Fourteen) address the role of demographic factors.
The penultimate chapter presents a number of case studies from the empirical data set
(Chapter Fifteen) The final chapter of Section C and the entire thesis (Chapter Sixteen)
reflects on the results of both the meta-analyses and the empirical study in terms of what
they reveal about the effects of the Holocaust on survivors and descendants. The fully
revised and tested version of the Model of the Differential Impact of Holocaust Trauma
across Three Generations is also presented.
© Janine Lurie-Beck 2007
16
Chapter Two – The Study of the Impact of the Holocaust on Survivors and their
Descendants
This chapter summarises the approach to the study of the psychological impact of the
Holocaust on survivors and descendants since the end of World War II.
Some
criticisms of this body of research are also presented.
2.1. – Impact on Survivors
The Holocaust is in our blood, in our bones and in our minds.
Kellerman (2006b)
You simply cannot get rid of that enduring pain – its always with you,
day and night.
A Holocaust Survivor
2.1.1. – History of the Assessment of Holocaust Survivor Mental Health
Formulations regarding the impact of the Holocaust on the human psyche were being
constructed even before survivors were liberated from their camps (Grubrich-Simitis,
1981; Levav, 1998). Eminent psychiatrists and psychologists such as Bruno Bettelheim
and Victor Frankl found themselves in an unenviable bird’s eye position while
themselves interned in concentration camps. Bettelheim and Frankl were among a
number of survivors who began to document not only their own experiences but
theorise as to the general psychological impact this historical event would have on its
survivors.
The thoughts of mental health workers dispatched to aid survivors in the
displaced persons camps soon appeared in print (for example Friedman, 1948;
Friedman, 1949; Niremberski, 1946). These early impressions paint a picture of initial
numbness and withdrawal, after which survivors appeared to revert back to a “normal”
level of functioning as they started to piece their lives together again (Niremberski,
1946). This appearance of normality may have been related to the happiness and relief
of being liberated. Friedman (1948) noted that untrained observers witnessing such
euphoric behaviour in the survivors they saw in the displaced persons camps mistakenly
saw it as a sign that they were unaffected by their experiences and would make a full
recovery. The assumption was that this post-liberation euphoria would reflect long term
prognosis.
The majority of initial aid efforts, however, were concentrated on survivors’
physical and material requirements (L. Berger, 1988; Bergmann & Jucovy, 1990;
Grubrich-Simitis, 1981). Medical attention was concentrated on organic injuries rather
© Janine Lurie-Beck 2007
17
than psychological ones (Eitinger, Krell, & Rieck, 1985; Krell, 1997a, 1997b;
Steinberg, 1989). Attention was focussed on the longitudinal impact of things such as
extreme starvation and infectious diseases such as typhus experienced by camp inmates.
Eitinger, Krell and Rieck (1985) note this was especially true for the publications
coming out of Europe during the 1950s and continued to be the main focus of literature
from Eastern European countries even up to the 1970s.
For almost a decade there was little attention paid to the psychological health of
Holocaust survivors (Gay, Fuchs, & Blittner, 1974).
Jucovy (1992) describes this
period as being drawn behind a “curtain of silence”. He argues that it seemed necessary
for survivors and the rest of the world to try to forget and move on (Jucovy, 1992). It
was as if the world wanted to forget about the horrific things that had happened
(Hodgkins & Douglass, 1984) or at least didn’t want to talk about it leading to a
“conspiracy of silence” (Danieli, 1998).
Krell (1997a) notes that psychological
interviews were rare in the immediate post-war period, and suggests that interviews at
this time were mainly conducted to gain information for physical assistance, relocation
or for war crime trials.
However, it needs to be remembered that very few survivors sought
psychological help in this period either. The blame for the lack of research in the first
post-war decade cannot therefore be solely placed on a lack of initiative on the part of
the scientific/psychiatric community. The fact that few survivors presented with mental
health issues during this period fed into the belief that to be a survivor you had to be
resilient and would therefore bounce back easily (Friedman, 1948, 1949). Friedman
(1948; 1949) suggested that many held the view that the fact that a survivor had
survived was proof of “physical and psychological superiority” and contends that this
may be why survivors’ psychological health was largely ignored at first. However, as
Grubrich-Simitis (1981) points out, “physical survival was no guarantee for psychic
survival”.
Several reasons abound for why survivors did not seek psychological help in the
initial post-war period. One suggestion is that survivors did not want to acknowledge
any weakness and wanted to appear strong and resilient (Dasberg, 1987). This façade
was not consistent with seeing a therapist. Kellerman (2001a) suggested that any
insinuation that survivors suffered from psychological problems could serve to
stigmatise them even more than they were simply by being identified as a survivor.
A second explanation is that survivors did not in fact suffer any symptoms
during this period and were going through a symptom free phase or latency phase in
© Janine Lurie-Beck 2007
18
their post-war adaptation. A number of clinicians have noted that some survivors
experienced a latency period during which they were relatively symptom free (Dasberg,
1987; de Wind, 1968/1995; Kren, 1989). This period lasted anywhere from a number of
months to a number of years (Ryn, 1990). For example Klein, Beersheba, Zellermayer
and Shanan (1963) noted that symptoms only appeared after about three years after
emigration to Israel among approximately 70% of survivors they came into contact with
in clinical practice.
Newman (1979) suggested that for some this latency period
continued until the survivors started having children when they were confronted with
issues that reminded them of their experiences during the war such as the loss of loved
ones. For some this period lasted as much as fifteen years (de Wind, 1968/1995)
Friedman (1949) ascribes this period to elation at being liberated from captivity
or having to hide after which a period of symptomatology inevitably follows. A number
of others suggested that the survivors were engrossed in the task of re-establishing their
lives (Dasberg, 1987; Klein et al., 1963; Newman, 1979; Solkoff, 1992b). Another
suggestion has been that this latency period coincided with the time when survivors still
had hope that they would find relatives alive and that it ended with the realisation that
this would never happen (Newman, 1979; Rappaport, 1968; Solkoff, 1992b). Once the
survivors were relatively settled, everyday events, or events related to the Holocaust
such as the Adolf Eichmann trial, would act as reminders of their Holocaust experiences
and trigger the emergence of symptoms (Klein et al., 1963; Solkoff, 1992b).
Ryn (1990) suggested that survivors went through at least four phases of postliberation adjustment which can also be used to explain the seemingly symptom-free
period. The first phase is dominated by physical symptoms as it is the time during
which the survivor received medical treatment to aid their recovery from starvation and
any diseases they contracted during their internment (Klein et al., 1963; Kren, 1989).
The second phase begins when this physical recovery is complete. Ryn (1990) argues
that after their physical recovery survivors go through a period during which both
physical/somatic and psychological symptoms are not apparent.
After this latency
stage symptoms begin to appear with social problems related to both family and outside
contexts also being noted. During the next phases symptoms such as anxiety and
depression become more ingrained (Ryn, 1990).
The main impetus for the growth of interest in the psychological well being of
Holocaust survivors was the passing of a law by the West German government in 1956
granting restitution to victims of Nazi persecution (L. Berger, 1988; Blumenthal, 1981;
Grubrich-Simitis, 1981; Hodgkins & Douglass, 1984; Jucovy, 1992; Krell, 1997a; Last,
© Janine Lurie-Beck 2007
19
1989). The program became known as Wiedergutmachung, meaning “to make good
again” (Bergmann & Jucovy, 1990; Fogelman, 1988). A large number of survivors
began presenting to psychiatrists seeking evidence of a link between their war time
experiences and their post-war psychological problems which was required for a
successful compensation application (L. Berger, 1988; Bergmann & Jucovy, 1990;
Bistritz, 1988; Hodgkins & Douglass, 1984; Jucovy, 1989, 1992; Last, 1989). As a
result, a number of psychiatrists and psycho-analysts began publishing case studies of
Holocaust survivors based on these compensation assessments (Grubrich-Simitis, 1981;
Hodgkins & Douglass, 1984). As the emphasis was on finding evidence of impairment
for compensation claims, the dominant theme arising from these early case studies and
theoretical discourses was a negative one of severe debility.
The early literature regarding the psychological impact of the Holocaust was
therefore based on single case studies or amalgamations of case studies of patients or
compensation applicants (Antonovsky, Maoz, Dowty, & Wijsenbeer, 1971; Dasberg,
1987; E. Harari, 1995; Hodgkins & Douglass, 1984). Thus these early formulations are
based on a small number of, what could be argued, least well adjusted of the survivor
population as a whole since they are based on survivors who were driven to seek
psychiatric help or felt their level of symptoms warranted compensation (Antonovsky et
al., 1971; Dasberg, 1987; Steinberg, 1989).
The study of Holocaust survivors led to a dramatic shift in thinking about the
source of psychological disturbance.
Up to this point the dominant theory (in
psychoanalytic circles at least) was that disturbances arose from personality problems
and failure to resolve developmental issues (Krell, 1997b; Z. Solomon, 1995) otherwise
organic brain damage had to be present (Grubrich-Simitis, 1981). That there could be a
lasting psychological impact of trauma was largely unacknowledged (Dasberg, 2001;
Krell, 1997a, 1997b). Unless a trauma caused physical injury and obvious damage to
the brain, even an extremely traumatic experience such as the Holocaust, was assumed
to result in only very temporary emotional problems which would disappear with the
cessation of the trauma itself (Grubrich-Simitis, 1981).
That psychological distress would be temporary is reflected in the terminology
used to describe the initial presenting symptoms of survivors immediately after
liberation such as “refugee neurosis” and “repatriation neurosis” (Grubrich-Simitis,
1981). These terms imply that these symptoms would dissipate once the survivor
became settled into life again.
© Janine Lurie-Beck 2007
20
Holocaust survivors presented many examples of clear psychological problems
not associated with obvious organic damage or childhood developmental arrest.
Psychiatrists were forced to acknowledge that psychological distress could be caused by
a traumatic event or series of events such as the Holocaust (Last, 1989). Chodoff
(1963) stated “when one considers the intensity of the stress undergone by these
patients, there seems little necessity to postulate any pre-existing personality
weaknesses or predisposition.” Niremberski (1946) stated “the causal agents of the
concentration camp mentality are of course obvious and entirely due to fear…and the
fight to exist”. This shift in thought led to theorising about a survivor syndrome which
became the forerunner to the post traumatic stress disorder diagnosis.
However the German officials considering the compensation claims remained
sceptical about the idea that psychological impairment could purely be the result of
living through an experience such as the Holocaust. An argument of a link between a
survivor’s symptoms and their Holocaust experience was much more readily accepted if
it could be linked with physical strains such as starvation or some form of physical
injury (Last, 1989). While some survivors were willing to endure the indignity of often
sceptical German officials questioning their claims for compensation, others chose to
ignore any psychological damage to avoid being further stigmatised by being identified
as being mentally ill (Fogelman, 1988).
Literature concerned with psychological impacts began to appear in larger
quantities from Western Europe, Canada and the United States and Israel in the 1960s
and 1970s (Eitinger et al., 1985; Krell, 1997b). Further opportunities to apply for
compensation have arisen in recent years via various European governments as well as
the possible access to Jewish monies from the war period held in Swiss banks (Brandler,
2000). This has led to a second wave of psychological assessments for compensation
applications and has increased interest in survivors’ current well-being (Brandler, 2000).
Clinicians soon noticed a pattern in the symptoms experienced by survivors and
as a result literature referring to a “syndrome” suffered by survivors, whether it was
termed the concentration camp syndrome as first described by Herman and Thyygesen
in 1954 (Brom et al., 2002), or KZ syndrome (Klein et al., 1963) or survivor syndrome
(Niederland, 1981, 1988). These syndromes which came to be applied to all survivors
were based on a clinical minority (Antonovsky et al., 1971). Conn, Clarke and Reekum
(2000) suggest that the creation of the survivor syndrome label had the effect of drawing
more attention to the plight of survivors. Once established as a recognised diagnosis it
also aided many in their efforts to seek compensation from the German government
© Janine Lurie-Beck 2007
21
(Conn et al., 2000). However, Berger (1988) suggests that clinicians soon became too
hasty in applying the diagnosis to any patient who was also a Holocaust survivor.
Depression, anxiety and paranoia are commonly listed as part of the
constellation of psychological symptoms seen in Holocaust survivors. In their seminal
work in the trauma field, McCann and Pearlman (1990) state that these three symptoms
are common reactions to many forms of trauma (p. 40).
2.1.2. – Depression
Depression is widely recognised as one of the most common symptoms experienced by
Holocaust survivors in the post-war period (Axelrod, Schnipper, & Rau, 1980; L.
Berger, 1988; Bergmann & Jucovy, 1990; Bistritz, 1988; Bower, 1994; Brom et al.,
2002; Chodoff, 1997; Dasberg, 2001; S. Davidson, 1980a; de Wind, 1968/1995;
Eitinger, 1973; Goldwasser, 1986; Grubrich-Simitis, 1981; Hafner, 1968; Hodgkins &
Douglass, 1984; Kellerman, 2001a; Klein et al., 1963; Krell, 1997b; Krystal, 1995; Last,
1989; Maller, 1964; Nathan, Eitinger, & Winnik, 1963; Niederland, 1981, 1988;
Niremberski, 1946; Porter, 1981; Rosenbloom, 1988; Ryn, 1990; Solkoff, 1981;
Steinberg, 1989). Porter (1981) argues that the severity of depression is strongly related
to the degree of survivor guilt experienced by the survivor. This survivor guilt, he
portends, relates to the loss of loved ones during the Holocaust (Porter, 1981).
Depression has also been considered to be a consequence of unresolved mourning for
the relatives that perished during the Holocaust (Solkoff, 1981; Steinberg, 1989).
According to Niederland (1981) depression is often masked by psychosomatic
symptoms.
2.1.3. – Anxiety
Along with depression, anxiety has been cited as one of the most lasting imprints left by
the Holocaust on a survivor’s psyche (Axelrod et al., 1980; L. Berger, 1988; Bergmann
& Jucovy, 1990; Bistritz, 1988; Bower, 1994; Brom et al., 2002; Chodoff, 1997;
Dasberg, 2001; S. Davidson, 1980a; de Wind, 1968/1995; Eitinger, 1973; Goldwasser,
1986; Grubrich-Simitis, 1981; Hafner, 1968; Hodgkins & Douglass, 1984; Kellerman,
2001a; Krell, 1997b; Krystal, 1995; Last, 1989; Maller, 1964; Nathan et al., 1963;
Niederland, 1981, 1988; Niremberski, 1946; Porter, 1981; Rosenbloom, 1988; Ryn,
1990; Solkoff, 1981). Anxiety displayed by survivors is often associated with the fear
of renewed persecution (Porter, 1981). Maller (1964) asserts that anxiety related to this
fear eventually evolves into a more generalised anxiety about contact with the outside
world. Contact with the outside world becomes a phobic fear to the point that the
survivor becomes almost agoraphobic (Maller, 1964).
© Janine Lurie-Beck 2007
Niederland (1981) describes
22
how someone knocking on the door or the phone ringing can arouse memories of when
the Gestapo arrived to forcibly remove them from their homes or from the ghetto and
therefore elicit anxiety reactions. Contact with anyone in uniform such as police can
also cause great anxiety (Niederland, 1981). These types of symptoms can be related to
those experienced by Posttraumatic Stress Disorder sufferers.
2.1.4. – Posttraumatic Stress Disorder Symptoms
The diagnosis of Posttraumatic Stress Disorder (PTSD) which first appeared in the
DSM-III (Wilson, 1986) grew out of the aforementioned survivor syndrome
formulations based on Holocaust survivors’ experiences (Ruedenberg-Wright, 1997). A
diagnosis of PTSD is applied when the presence of a threshold combination of three
symptom clusters of intrusion, avoidance and hyperarousal is reported/observed.
Intrusion is defined in the DSM-IV-TR as persistent re-experiencing of the traumatic
event/s such as via dreams, images, thoughts or flashbacks. Avoidance is defined as
persistent avoidance of stimuli associated with the trauma and numbing of general
responsiveness for example avoiding activities, places or people that act as reminders of
the trauma. Hyperarousal is defined as persistent symptoms of increased arousal as
evidenced by symptoms such as difficulties sleeping and/or concentrating,
hypervigilance and exaggerated startle response.
The quintessential PTSD symptoms of intrusion and avoidance have been
recognised among survivors (for example Favaro, Rodella, Colombo, & Santonastaso,
1999; Lavie & Kaminer, 1996; Lev-Wiesel & Amir, 2003; Yehuda, Kahana, Southwick,
& Giller Jr, 1994). Survivors are often plagued by intrusive dreams and flashbacks of
their Holocaust experiences and go to great lengths to avoid contact with things that
may trigger such intrusions. Such symptoms are commonly associated with co-morbid
anxiety and depression problems (McFarlane & Yehuda, 1996) and have been seen to
do so for the Holocaust survivor population as well (Favaro et al., 1999; Yehuda et al.,
1994).
2.1.5. – Paranoia/Fear of Further Persecution
Paranoia is another psychological variable commonly linked with Holocaust survival
(Axelrod et al., 1980; L. Berger, 1988; Bergmann & Jucovy, 1990; Bistritz, 1988;
Bower, 1994; Chodoff, 1997; S. Davidson, 1980a; de Wind, 1968/1995; Goldwasser,
1986; Kellerman, 2001a; Krell, 1997b; Last, 1989; Maller, 1964; Niederland, 1988;
Niremberski, 1946; Solkoff, 1981). This paranoia is said to be an escalation of anxieties
related to the fear of renewed persecution.
Those affected appear “chronically
apprehensive and afraid to be alone” (Niederland, 1988). Davidson (1980a) suggests
© Janine Lurie-Beck 2007
23
that a tendency towards suspiciousness is sometimes the most obvious symptom
experienced by survivors.
2.1.6. – Interpersonal Trust and Intimacy
Another area of functioning mentioned by a great number of theorists/researchers is
survivors’ ability to develop trust and intimacy in personal relationships. It has been
contended that because of their experiences survivors are impaired in their ability to
form secure attachments (L. Berger, 1988; Brom et al., 2002; E. Cohen, Dekel, &
Solomon, 2002; Dasberg, 2001; S. Davidson, 1980a; Eitinger, 1973; Freyberg, 1980;
Grubrich-Simitis, 1981; Kellerman, 1999, 2001a; Klein et al., 1963; Kren, 1989; Nadler
& Ben-Shushan, 1989; Nathan et al., 1963; Porter, 1981; Rosenbloom, 1988; Ryn,
1990).
It has been suggested that this “insecurity in human relations” developed as a
consequence of Holocaust trauma was the most important and detrimental effect of
survival (de Wind, 1968/1995). This is supposed to have occurred because of the
sudden and often brutal way in which survivors were separated from their parents,
spouses, children and/or other relatives (E. Cohen et al., 2002; Freyberg, 1980; Prot,
2000). Davidson (1980a) suggested that this resulted from the fear of the pain of further
object loss.
They had already lost so many loved ones that they dare not allow
themselves to get close to anyone else because it would hurt them too much if they lost
them.
Difficulties in developing trust in relationships is a common reaction to any
traumatic event among all age groups (Macksoud, Dyregrov, & Raundalen, 1993).
McCann and Pearlman (1990) note that damaged trust schemas among people who have
been victimised is a common occurrence (p. 44).
2.1.7 – Factors Affecting the Severity of the Impact of Holocaust Trauma
2.1.7.1. – Coping styles and strategies.
Research into other potentially key psychological factors has been largely lacking. In
particular, coping skills/strategies have been cited as a very important determinant of
post-traumatic adjustment and have been related to symptom levels and attachment
dimensions (for example Aronoff, Stollack, & Sanford, 1998; Carver, Scheier, &
Weintraub, 1989; Fogelman & Savran, 1979; B. L. Green et al., 1985; McFarlane &
Yehuda, 1996; Stone, no date; Wilson, 1989). Coping strategies/resources are included
in both Green et al.’s (1985) Working Model for the Processing of a Catastrophic Event
and Wilson’s (1989) Person-Environment Approach to Traumatic Stress Reactions as
important determinants of psychological impact of traumatic events. There has been
© Janine Lurie-Beck 2007
24
limited research into how well survivors cope with subsequent traumas such as a cancer
diagnosis, but little research could be located that simply addressed coping in everyday
life.
2.1.7.2. – World assumptions.
While
there
has
been
a
reasonable
amount
of
research
conducted
into
psychopathological symptoms/disorders such as depression and anxiety among
survivors, a potential underlying mechanism for the emergence of these symptoms
among survivors has been under researched. What is referred to here is the concept of
world assumptions as espoused by Janoff-Bulman (1992) in the Theory of Shattered
Assumptions.
The scale and over-whelming nature of the Holocaust meant that for many
survivors their concept of the world changed forever. Janoff-Bulman (1992) argues that
we all have the central beliefs that the world is a relatively fair and just place. Such
beliefs were often shattered by the Holocaust. The extent to which this occurred had
the potential to influence the degree of symptoms experienced by survivors in the postwar period (McFarlane & Yehuda, 1996; Valent, 1995).
McCann and Pearlman (1990,
p. 32) explain how traumatic memories are often completely incongruous with the
survivors’ pre-existing schemas about the world. This incongruity often leads to the
development of symptoms such as depression, anxiety and more specifically PTSD
symptoms such as avoidance (McCann & Pearlman, 1990, p. 51). The shattering of
these assumptions is more marked when the trauma is human-induced as was the case
for the Holocaust – the survivor has to acknowledge that their trauma was caused by
other human beings (Janoff-Bulman, 1992, p 78).
A relationship between world
assumptions and symptom levels has been found in at least one study with Holocaust
survivors (Brom et al., 2002). The following quote gives an indication of how a
survivor’s assumptions about the world could be so affected by their Holocaust
experiences.
And when you think of it what a man can do to a man the only way to
be sane you have to believe that they were all insane the Germans
because if they weren’t insane you wouldn’t want to live in the world
would you. Sometimes when the sun is shining and I look around
and the birds are singing and the flowers are blooming I think to
myself that is not possible, it just didn’t happen, it couldn’t have, it
couldn’t have. But it did. But it did.
Jozefa Lurie, a Polish Jewish camp survivor
© Janine Lurie-Beck 2007
25
2.1.8. – Posttraumatic Growth
Another more recent area of research within the trauma field has been that of posttraumatic growth. Tedeschi (1999) defines posttraumatic growth as “important changes
in perception of self, philosophy of life, and relationships with others in the aftermath of
events that are considered traumatic in the extreme” (p.321).
Living through a
traumatic event can lead to positive changes in a person’s perception of their ability to
deal with difficult circumstances. With the benefit of hindsight, the survivor reflects
that they were quite capable in the way they coped with the traumatic experience which
insight then leads to a more positive perception of their own abilities and may also
positively influence they way they deal with later experiences (Tedeschi & Calhoun,
1996). In terms of philosophical changes, a survivor or victim of trauma may become
more appreciative of life, vow to life live to the fullest and may (after a possible
weakening of beliefs) be left with stronger spiritual and religious beliefs, related to the
knowledge that they faced adversity and lived through it (Tedeschi & Calhoun, 1996).
Finally in the area of relationships with others, Tedeschi and Calhoun (1996) contend
that in their post-trauma life, survivors or victims may value their relationships more,
and also develop deeper and more intimate relationships, partially as a result of
sharing/discussing their traumatic experience with loved ones and friends.
Only one study was located that addressed post-traumatic growth among
survivors. It is therefore of interest to further explore this issue in relation to Holocaust
survivors. Janoff-Bulman (1992, p 136) states that the experience of a traumatic event
as large in scale and effect as the Holocaust often leads the survivor to reassess their
values and beliefs in relation to themselves and what they consider important in life and
in the world.
It would seem intuitive to predict that posttraumatic growth would be inversely
related to negative traumatic impacts but the balance of research in this area tends to
suggest that they co-exist. Both the study conducted with survivors (Lev-Wiesel &
Amir, 2003) as well as others with other population groups such as carers of AIDS
patients (Cadell, Regehr, & Hemsworth, 2003), Israeli adolescents exposed to terror
incidents (Laufer & Solomon, 2006), brain-injury survivors (McGrath & Linley, 2006)
and general population members who have suffered varying traumatic incidents
(Morris, Shakespeare-Finch, Rieck, & Newbery, 2005) obtained positive relationships
between negative symptomatology and posttraumatic growth. The finding of a positive
relationship between posttraumatic growth and pathological symptoms is in contrast to
what is implied in Green et al.’s (1985) trauma model (Processing a Traumatic Event:
© Janine Lurie-Beck 2007
26
A Working Model). In their model, Green et al. (1985) denote post-trauma adaptation
as characterised either by growth/restabilisation or pathological outcome. It is granted
that Green et al. (1985) formulated their model before much of the research on
posttraumatic growth was conducted, but it is still interesting to note that they seem to
suggest a mutual exclusivity between pathological symptoms and posttraumatic growth.
The relationship between posttraumatic growth and negative symptomatology is
therefore worthy of further exploration within the context of Holocaust survival and this
is conducted within the current thesis.
2.2. – Impact on Children of Survivors
I can see that my whole adult life has been completely driven by what
happened to my family during the Holocaust.
A child of survivors
The Holocaust is always there in the background.
A child of survivors
The first publications regarding the impact of the Holocaust on children of survivors
came from Canada in the mid 1960s (Axelrod et al., 1980; Bergmann & Jucovy, 1990;
Brom, Kfir, & Dasberg, 2001; S. Davidson, 1980a; Grubrich-Simitis, 1981; Last, 1989;
Newman, 1979; Vogel, 1994). This was as a result of a notable increase in the number
of children of survivors presenting for therapy (Brom et al., 2001; S. Davidson, 1980a;
Grubrich-Simitis, 1981; Newman, 1979). Kestenberg and Kestenberg (1990a) add that
it was around this time that a possible connection between symptoms and parents’
survivor status was recognised. Children of survivors had been seen before this but the
fact that their parent or parents were Holocaust survivors had largely been ignored
(Kestenberg & Kestenberg, 1990a).
Researchers and clinicians noted that children of survivors displayed symptoms
reminiscent of those found among survivors (Gay et al., 1974; Grubrich-Simitis, 1981;
Kellerman, 2001a; Rowland-Klein & Dunlop, 1997; Ryn, 1990; Steinberg, 1989;
Zilberfein, 1996).
Some noted that the children presented as if they had actually
experienced the Holocaust first hand themselves (Yehuda, Schmeidler, Wainberg,
Binder-Brynes, & Duvdevani, 1998; Zilberfein, 1996).
The 1960s coincided with a time when a large number of children of survivors
were reaching adolescence and young adulthood (S. Davidson, 1980a; Eitinger et al.,
1985; Newman, 1979; Russell, 1982). Adolescence was a particularly common point
of crisis (Newman, 1979). The first paper examining children of survivors, written by
© Janine Lurie-Beck 2007
27
Rakoff, Sigal and Epstein (1966) appeared at a time when children born immediately
after the war were on the cusp of adolescence and young adulthood.
Again, depression, anxiety, paranoia and intimacy/trust issues are among the
most common symptoms associated with children of survivors (Zilberfein, 1996).
Problems establishing relationships because of these intimacy/trust issues have been
found to be one of the biggest issues for these children (Barocas & Barocas, 1980; S.
Davidson, 1980a; Fogelman & Savran, 1979; Freyberg, 1980; Kellerman, 1999;
Kuperstein, 1981; Reijzer, 1995; Rieck, 1994; Russell, 1982; Shoshan, 1989; Zilberfein,
1996). In fact it was precisely these relational factors that drove some of this group to
seek psychological assistance (S. Davidson, 1980a; Rieck, 1994).
Elevated anxiety levels have been frequently related to children of survivors.
(Felsen, 1998; Freyberg, 1980; Gay et al., 1974; Grubrich-Simitis, 1981; Kellerman,
1999, 2001a; Kuperstein, 1981; Rowland-Klein & Dunlop, 1997; Solkoff, 1992b;
Steinberg, 1989; Yehuda, Schmeidler, Giller, Siever, & Binder-Brynes, 1998;
Zilberfein, 1996). This anxiety has been related to the pressure of meeting unrealistic
expectations held by survivor parents (for example Kuperstein, 1981).
Kuperstein
(1981) notes that these children often suffer from examination anxiety as a result of
pressure to do well in their studies. Anxiety has also been linked to survivor parents’
tendency towards being overprotective and portraying the world as a dangerous place
(Rowland-Klein & Dunlop, 1997).
Paranoia has also been listed as a symptom
frequently experienced by children of survivors (S. Davidson, 1980a; Rustin, 1988;
Solkoff, 1992b).
Depression
also
dominates
descriptions
of
children
of
survivors
symptomatology (S. Davidson, 1980a; Felsen, 1998; Freyberg, 1980; Gay et al., 1974;
Grubrich-Simitis, 1981; Kellerman, 1999, 2001a; Kuperstein, 1981; Newman, 1979;
Perel & Saul, 1989; Rowland-Klein & Dunlop, 1997; Russell, 1982; Rustin, 1988;
Solkoff, 1992b; Steinberg, 1989; Yehuda, Schmeidler, Giller et al., 1998). It has been
argued that children of survivors’ depression is in actual fact anger turned inwards
(Newman, 1979; Rowland-Klein & Dunlop, 1997). They feel anger towards their
parents but are unable or unwilling to direct it to their parents (Rowland-Klein &
Dunlop, 1997). Depression has also been linked to children’s guilt feelings about
wanting to leave home: another issue associated with adolescence (Grubrich-Simitis,
1981; Russell, 1982). Guilt feelings associated with the knowledge of their parents’
own suffering mean that children feel unable to rebel against their parents or fight the
© Janine Lurie-Beck 2007
28
aforementioned unrealistic expectations and so become depressed as a result (Newman,
1979; Perel & Saul, 1989).
As with the literature relating to Holocaust survivors themselves, early clinical
research tended to paint a more pessimistic picture of children of survivors than later
non-clinical research (Solkoff, 1992b; Yehuda, Schmeidler, Giller et al., 1998).
Therefore some literature suggests that children of survivors have higher incidence
levels of psychological symptoms than those with non-survivor parents while other
literature suggests they are no different to the general population.
There is disagreement about the proportion of children of survivors with
emotional problems.
Davidson (1980a) states that up to one fifth of the children
referred to child psychiatric and adolescent outpatient services in Israel had at least one
survivor parent. However, he fails to note whether this proportion is actually higher than
the proportion of the Israeli population with survivor parents. Despite this Davidson
(1980a) argues that there is a high incidence of psychological problems among the
children of survivor population. In contrast, Gay et al. (1974) state that children of
survivors did not present to their clinic (also in Israel) in higher proportions than were
found in the population.
2.3. – Impact on Grandchildren of Survivors
Being the granddaughter of survivors has always influenced my
identity.
A grandchild of survivors
I often feel like the Holocaust is not something I remember because
someone told me about it, but rather, something I remember because I
was there.
A grandchild of survivors
Assessment of grandchildren of Holocaust survivors has slowly been building up over
recent years. For a while many have theorised that effects would be evident in this third
generation but it has only been recently that this group have become young adults in
large enough numbers to be more accessible as a research group.
Bistritz (1988) suggested that there was evidence for pathology among the
grandchild of survivor group. Ryn (1990) noted two years later that reports of
disturbances among grandchildren of survivors were beginning to appear.
However, as Chaitin (2003) points out, there is still much to be learned about
grandchildren of survivors.
Because of the lack of research to date there are no
conclusive findings as to the mental health of this group, only theoretical conjecture
based largely on anecdotal evidence. However, Ryn (1990) did suggest that there was
© Janine Lurie-Beck 2007
29
some evidence of a “sui generis” inheritance of the disturbances suffered by survivors
and their children. In other words, similar types of symptoms seemed to be appearing in
the third generation.
2.4. – Summary of the Impacts of the Holocaust across Three Generations
This chapter so far has summarised the symptomatology suffered by Holocaust
survivors and their descendants. The psychological symptoms and attributes that have
been discussed in this chapter have been evident among survivors and their descendants.
These psychological symptoms and attributes have been presented diagrammatically in
Figure 2.1. Figure 2.1 represents the first stage of a preliminary Model of the
Differential Impact of the Holocaust across Three Generations which will be expanded
on in the coming chapters of this section of the current thesis (Section A). The two
variables of world assumptions (assumptions that the world is a fair and predictable
place) and coping strategies that have been suggested as possible influential factors in
the degree to which the psychological effects are noted in each of the generations have
been placed in this model as influential psychological processes. The degree to which
world assumptions have been affected and the usage (or lack of usage) of negative and
positive coping strategies are suggested as being at least partially predictive of the
severity of psychological impact of the Holocaust.
Psychological Impacts of the
Holocaust
Influential Psychological
Processes
1st Generation
(Survivors)
Depression
Anxiety
Paranoia
PTSD symptoms
Romantic Attachment
Dimensions
• Post-traumatic Growth
• World Assumptions
• Coping Strategies
2nd Generation
(Children of
Survivors)
•
•
•
•
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
• World Assumptions
• Coping Strategies
3rd Generation
(Grand-children
of Survivors)
•
•
•
•
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
• World Assumptions
• Coping Strategies
•
•
•
•
•
Figure 2.1. Preliminary Model of the Differential Impact of Holocaust Trauma on Three Generations
© Janine Lurie-Beck 2007
30
2.5. – Critique of Research regarding Holocaust Survivors and Descendants
Methodologies utilised in the study of Holocaust survivors and their descendants to date
have been inadequate on a number of levels. Some of these problems were recognised
many years ago and yet the implementation of recommendations for improved methods
has not been widespread.
Problems relate to the definition of what constitutes a
“Holocaust survivor” or a “Child of a Holocaust survivor”, appropriateness of control
groups, differences between results based on clinical and non-clinical samples and lack
of the inclusion of the assessment of demographic factors. Each of these problems will
be addressed below.
2.5.1. – Definition of “Holocaust Survivor” and “Child of a Holocaust Survivor”
Survivors of the Holocaust endured many different types of trauma ranging from
persecution to ghettoisation and concentration camp internment. Researchers differ as
to which of these experiences should be included in a Holocaust survivor categorisation.
Some researchers reserve the term Holocaust survivor solely for those who were
interned in a concentration camp for a period during the war (Lev-Wiesel & Amir,
2000). Others also include those who managed to escape camp internment by living in
hiding and those who survived under the protective umbrella of partisan groups and the
resistance (Felsen, 1998; Hodgkins & Douglass, 1984). Felsen (1998) also located
studies that broaden the definition to include anyone who was domiciled in any region
of Europe under Nazi occupation during the war, with some arguing that anyone who
survived the war in any part of Europe or the USSR ought to be included.
The operationalisation of a “child of Holocaust survivors” has also been fraught
with inconsistencies. Some studies confine this category to those with two survivor
parents while others also include those with only one survivor parent (Felsen, 1998). A
few studies were found which included participants born during the war (for example
Gertz, 1986). Sigal and Weinfeld (1987) and Kestenberg and Kestenberg (1990a)
duly note that any effects noted in these children could have originated from their own
exposure to Holocaust trauma as opposed to influences from their parents.
The differences in definition of Holocaust survivor also comes in to play as
researchers define “child of a Holocaust survivor” by the experiences of the parent.
Therefore inconsistencies in this definition also flow through to studies of the next
generation.
Discrepancies in the definition of the target population obviously have
implications for the comparability of study results. The reader must bear in mind the
criteria used for inclusion in the study group when considering results. In the research
© Janine Lurie-Beck 2007
31
discussed in the current thesis, a broad definition of survivors and their descendants is
used, with data collected to identify sub-groups within these populations.
Specifically:
• A survivor is defined as a person who endured some form of persecution by the Nazi
regime. This was dated from January 1933 (when Hitler came to power) until the
end of World War II.
• A child of survivor/s is defined as having at least one parent meeting the survivor
criterion and a birth date after 1945 or after their survivor parent’s persecution had
ended (in the case of survivors who escaped Europe prior to 1945).
• A grandchild is further defined by having at least one grandparent meeting the
survivor criterion and at least one parent meeting the child of survivor/s criteria (and
neither parent meeting the survivor criteria).
2.5.2. – Nature of Control Groups
The predominant control group used in studies of Holocaust survivors is made up of
people who emigrated from Europe (specifically countries that were under Nazi
occupation) before the war (i.e., pre 1939). Other studies merely used people native to
the country the study was conducted in as a control group. There are also a number of
studies which fail to even specify the nature of their control group at all (Lomranz,
1995).
The argument that aspects of Holocaust survivors post-war adjustment (and also
their children’s well being) could be related to the experience of migrating to a new
country, rather than their traumatic war experiences, gained currency in the last two
decades (Baranowsky, Young, Johnson-Douglas, Williams-Keeler, & McCarrey, 1998;
Halik, Rosenthal, & Pattison, 1990; Kuperstein, 1981; Solkoff, 1992b; E. Weiss,
O'Connell, & Siiter, 1986). In light of this, the use of native born subjects as a control
group (or as a sole control group) has fallen from favour. The vast majority of studies
examined for this thesis used pre-war immigrants as their control group.
However, the use of pre-war immigrants as a control group also potentially
introduces confounds to research. Shmotkin and Lomranz (1998) assert that there are
two issues that have rarely been acknowledged by researchers utilising this type of
control group. Probably the vast majority of these pre-war immigrants left behind some
family members in Europe. These family members could potentially have died during
the Holocaust. As a case in point, in a study by Bistritz (1988) 40% of the group
considered to be a control group had lost family members in the Holocaust. Therefore,
while they did not endure the Holocaust themselves they had to mourn the loss of
© Janine Lurie-Beck 2007
32
family members who did (Kahana, Harel, & Kahana, 1989). The impact of losing
family members in such tragic circumstances has been suggested as one of, if not the,
most traumatising aspects of Holocaust survival (see Chapter Four for more details).
Illustrating this point is Nathan, Eitinger and Winnik’s (1964) finding that pre-war
immigrants whose parents died in the Holocaust had similar levels of psychological
symptoms to a survivor group, while both groups evidenced more symptoms than a prewar immigrant group whose parents had not died during the Holocaust. These people
can hardly be considered untouched by the Holocaust as is assumed by their inclusion in
a “control group”.
The second point raised by Shmotkin and Lomranz (1998) is that while the use
of pre-war immigrants as a control group is meant to control for the immigration effect
noted earlier, the nature of the pre- and post-war immigration experiences differed
markedly. Post-war immigrants often had to endure extended stays in displaced persons
camps before their chosen settlement country allowed them to emigrate. Competition
for such placements was much fiercer in the post-war period as there were so many
more people wanting to emigrate.
Kahana et al. (1989) have suggested that the best control group to use, which
eliminates the aforementioned problems with differences in pre and post war
immigration experiences, is one consisting of people who lived in Europe during the
war but experienced minimal if any persecution. They suggest the ideal group would be
post-war Romanian immigrants as Romania remained relatively untouched until the
very late stages of the war.
Kahana et al. (1989) also suggested that multiple control groups might be used
in order to eliminate/control for any potential confounding variables that have been
discussed. Only two studies that utilised more than one control group were found by
the current author in her literature search for this current thesis.
Joffe, Brodaty,
Luscombe and Ehrlich (2003) had a pre-war immigrant control group as well as an
Australian or English born control group in their study conducted in Sydney. The
inclusion of both groups allowed an examination of the potential immigrant effect. The
pre-war immigrant group was found to score higher on both depression measures than
the native born control group providing support for the immigration effect. In fact, the
pre-war immigrant group scored statistically significantly higher on the measure of
severe depression (t (98) = 2.60, p < 0.05). However, both control groups obtained a
mean of 0.80 on the measure of anxiety used (Anxiety and Insomnia scale on the
General Health Questionnaire); a result that does not lend support to the idea that
© Janine Lurie-Beck 2007
33
immigrant status increases psychopathology symptoms.
Shmotkin, Blumstein and
Modan (2003) had both a pre-war and a post-war immigrant control group. This
allowed an assessment of the impact of differential immigrant experiences with postwar immigrants presumably experiencing similar immigrant experiences to the
Holocaust survivors themselves. The two immigrant groups scored almost identically
on the depression measure used.
Carmil and Carel (1986) suggest that studies conducted with survivors who live
in Israel should include a control group from another country. They argue that given
the various wars that Israel has been involved in since its inception in 1948 the
population as a whole, irrespective of survivor status, may have a higher level of
psychological distress.
In the current thesis, the nature of control groups used is elucidated and
discussed. The potential immigration effect is also examined in the empirical study
where comparisons are made between children of survivors born before and after their
survivor parents’ emigration.
2.5.3. – Sample Recruitment Methods
While some researchers seek volunteers via appeals to the general public, the majority
make contact with potential subjects via membership lists of organisations.
Such
organisations are either generic Jewish organisations which may include both survivor
and non-survivor members (Felsen, 1998) or organisations specifically for survivors.
The use of member lists such as these prompts the question of whether members are
different to non-members. Using membership lists from Jewish organisations could be
seen as the preferable option as it may include survivors who don’t necessarily want to
broadcast their survivor status and would therefore be less likely to join a survivor
group. However, bias is still present as only people who more strongly identify with
their Jewish identity and heritage are likely to join such groups (Levav, 1998). As
Levav (1998) argues, not all potential respondents will be contactable via such a group.
This method will also only derive Jewish survivors and will miss out on contacting nonJewish survivors.
The derivation of samples via appeals to survivor group members however is
even more limiting. This method only contacts survivors who have joined such groups
and therefore were willing to identify themselves as a survivor. Differences between
members and non-members of survivor groups as well as the reasons why a survivor or
child of a survivor may want to join a support group of this nature have been discussed
by a number of researchers.
© Janine Lurie-Beck 2007
Some argue that membership is beneficial for the
34
psychological well-being of survivors while others contend that they are a magnet for
the less well-adjusted among the survivor group.
Fogelman and Savran (1979), referring to children of survivors, argue that those
who don’t feel they have been particularly impacted upon by their parent/s experiences
are far less likely to join a group. In their experience, it is the children who are
struggling to come to terms with their parent/s experiences and what they mean to them
that join such groups in an attempt to resolve such issues. This would mean that
samples derived from such groups are starting off from a more affected sub-group of the
population.
Baron, Reznikoff and Glenwick (1993) contend that the literature they examined
suggests that participation in such support groups has a positive influence and leads to a
decrease in level of symptoms experienced. The survivor derives comfort from the idea
that they are not unique and that others share similar experiences. Kahana and Kahana
(2001) state that the group serves as a substitute extended family for the survivor and
the presence of this extra support network therefore reduced the need for more formal
psychological intervention. As mentioned in Chapter One, Section 1.1.4, survivor
groups based around survivor’s pre-war region or town served to ameliorate some of the
negative feelings associated with the abrupt removal from pre-war communities (Zolno
& Basch, 2000)
Gertz’s (1986) qualitative appraisal of reasons for membership among a group
of children of survivors revealed that issues of identity were central. Members stated
that they joined a survivor group because they strongly identified with their role as the
child of a survivor and wanted to acknowledge that. They further developed this
identity by contact with others in the same position which they met through the group.
For many children of survivors, group membership serves to provide a surrogate
extended family that was not available to them within their own family because of the
death of grandparents during the Holocaust. Learning that other children of survivors
had similar experiences with their survivor parents as they were growing up also adds to
the strong sense of camaraderie that develops in such groups (Zolno & Basch, 2000).
Robinson, Rapaport-Bar-Sever and Rapaport (1994) found that membership of
survivor organisations among survivors themselves is related to demographic variables
such as age of the survivor and the nature of their Holocaust experiences. They found
that survivors who had spent time in hiding were less likely to have joined this type of
support group than survivors who had been in a camp. They also found that the
likelihood of joining a group increased with age. That is, the older the survivor was
© Janine Lurie-Beck 2007
35
when their persecution began the more likely they were to be a member of a survivor
support group.
Clearly, the issue of sampling recruitment methods is one which must be
considered when evaluating any research study involving Holocaust survivors or their
descendants.
Unfortunately, readers are denied the opportunity to deem for themselves
the effectiveness of sampling techniques when examining a fair number of publications
of survivor research. Solkoff, who has conducted several reviews of the literature in
this area (Solkoff, 1981, 1992a) found that samples are often not properly described in
terms of the sampling methods used to obtain them (Solkoff, 1992b). The current
author also found that a sizeable proportion of studies fail to give enough detail to
enable exact replication of sampling.
2.5.4. – Differences between Clinical and Non-clinical Study Results
While it is clear that some survivors have suffered a great deal in the post-war period
from a number of symptoms, others have managed to adapt quite well.
Studies
examining the incidence levels of symptoms do not find that 100% of those surveyed
suffer from these symptoms. For example, while Szymusik’s (1964 (in Polish), cited in
Ryn, 1990) assessment of a group of former camp inmates in Poland found that 60%
were considered to have some psychological problems this leaves 40% who did not.
Helweg-Larsen et al. (1949 (in Danish), cited in Nathan et al., 1964) found 75% of a
group of 130 camp survivors in Denmark displayed neurotic symptoms of varying
degrees of severity, leaving 25% relatively symptom free. Later studies have also found
a percentage of survivors citing no psychological distress; for example 26% of a sample
studied by Carmil and Carel (1986) and 39% of a sample studied by Chaitin (2002).
Further evidence that some survivors have adapted better than others comes
from inconsistencies in the literature between clinical and the non-clinical studies which
began to appear from the 1970s. Research based on non-clinical groups of Holocaust
survivors often find no or little difference between them and control groups whereas
clinical research tends to paint a bleaker picture (L. Berger, 1988; Gross, 1988; Rustin,
1988). The most parsimonious explanation for this is that clinical samples are made up
of survivors whose symptoms were debilitating enough for them to seek some form of
psychological help which presupposes that their post-war adjustment was not as good as
participants in non-clinical studies who mostly did not seek such help. Antonovsky et
al. (1971) suggest that it is inevitable that clinical samples will be found to evidence a
degree of psychopathology since “by definition, patients are maladaptive.” That such
results cannot, and should not, be applied to the entire Holocaust survivor population
© Janine Lurie-Beck 2007
36
(which includes a large number who did not find the need to seek psychiatric or
psychological help) is a moot point. Clinical sub-samples of the Holocaust survivor
population cannot be considered representative of the survivor population as a whole
(Dasberg, 1987; Whiteman, 1993; Yehuda, Schmeidler, Giller et al., 1998). There are
also some survivors who perhaps did suffer symptoms to a degree that might warrant
intervention or compensation but did not seek any. In a 1947 to 1951 study of 1,300
Danes who had been in camps about 75% stated they had had or still had neurotic
symptoms of varying degrees of severity (Eitinger, 1961). Very few of them had sought
psychiatric help. Therefore, on a number of levels, clinical studies cannot even be
considered representative of the clinical sub-population of survivors.
Berger (1988) suggests that because there are differences in results within the
literature (between studies based on clinical and non-clinical samples) “at least some of
(and possibly all) of the results are inaccurate”. This seems to the current author to be
the most illogical argument of all since it fails to take into account the notion that results
may differ purely because they are based on different samples of survivors who have
differing levels of symptomatology. As Krell, Suedfeld and Soriano (2004, p 505) put it
It is important not to use the lives of well-adjusted survivors as a basis
for neglecting the severe long-term effects of the Holocaust on some
and equally important not to commit the error of over-generalising
from those who seek therapy or other help to the entire population of
survivors.
Gross (1988) makes a valid point in regard to this issue that few seemed to have
thought about before. Initial publications in relation to post-war adaptation of survivors
were based on summaries of clinical case notes of survivors who had either sought
psychiatric assessment in order to apply for compensation or had sought therapy. Nonclinical survey research emerged in later years. The clinical literature tends to present a
more negative picture than the later community research. Gross (1988) points out that
the very fact that the clinical literature appeared earlier than the survey research could
be the key to why they differ in their conclusions. Non-clinical survey research was
conducted with survivors when a longer amount of time had passed between the
Holocaust and the study participation. Clinical case studies were published at a time
that was more contemporaneous. Therefore what we may be seeing here is merely an
alleviation of symptoms over time (Gross, 1988). Whiteman (1993) states that many of
the early formulations of the impact of the Holocaust on survivors continue to be taken
for granted today even though they are clearly based on evaluations conducted when
survivors were “at their lowest ebb”.
© Janine Lurie-Beck 2007
This has led to a tendency towards
37
“syndromisation” of survivors which has also been found to occur with children of
survivors as well (Steinberg, 1989).
When reviewing the literature many authors have suggested that there is no
overall difference between survivors (or their descendants) and the general population
because a large number of non-clinical studies find no statistically significant
differences between them and control groups. However, a vast majority of these studies
do find that survivor groups score higher on measures of mental health even if it is not
statistically significantly higher. The clinical groups have statistically significantly
higher levels psychopathological symptoms/disorders while the non-clinical groups
have non-statistically significantly higher levels than control groups/the general
population.
Felsen (1998) suggests that the term complex rather than syndrome be used to
reflect that the differences between survivor groups and controls are there to some
degree. The term syndrome implies that symptoms are present to a clinically significant
level while the term complex, Felsen (1998) argues, can be used to merely identify a
differing psychological profile with mild, but not necessarily pathological, levels of
certain symptoms.
Felsen (1998) applied this argument to children of survivors,
however it could be used to apply to any non-clinical group of children of survivors or
survivors themselves. So the key issue should be why some survivors end up being
classified as clinical and others non-clinical. What factors lead to these different levels
of post-war adjustment? Many of these factors will be demographic in nature.
2.5.5. – Assessment of Demographic Differentials
Perhaps the biggest criticism that can be levelled at research conducted with survivors
and their descendants, which also represents an unfortunate missed opportunity, is the
lack of assessment of demographic differences within the survivor population and their
descendants. The vast majority of research with this population has concentrated on
differences between this group as a whole compared to control groups (Kahana &
Kahana, 2001). As Fogelman and Savran (1979) noted, what was missing was an
indication of the range and degree to which symptoms were present in this population.
In other words how the levels of adaptation differed within the survivor group among
different demographic subgroups.
While efforts have been increasing in recent years, very few researchers have
examined differences between sub-groups as readily assessable as gender and age
(Solkoff, 1992b).
What makes this fact even more puzzling is the fact that
heterogeneity among the Holocaust survivors and their descendants has been argued
© Janine Lurie-Beck 2007
38
within theoretical discourses from the very beginning (Blank, 1996; Eitinger, 1969;
Niremberski, 1946; Yehuda, Schmeidler, Giller et al., 1998) despite some arguing that
heterogeneity was only recognised much later (Steinberg, 1989).
Only a very small minority of community survey studies conducted sub-group
analysis to check for these potential differences. Some studies fail to even provide a
description of the demographic characteristics of their sample (Solkoff, 1992a), let
alone using them as variables for further statistical analysis. Others have noted how
important it is to look for within group differences but then fail to report information
about this issue in relation to their own study (for example Blank, 1996).
In his reviews of the literature in this field Solkoff (1992b) points out the lack of
investigation of issues such as religious/ethnic backgrounds (he gives the example of
comparing Polish Jewish survivors to Polish Catholic survivors), nature of Holocaust
experiences (i.e., camp incarceration versus living in hiding), and country of post-war
settlement (for example Israel versus America or Australia versus remaining in Europe).
While a number of articles that examined the impact of differing Holocaust experiences
were found, the issues of ethnicity/religion and post-war settlement location were
ignored by the majority of researchers. Only one study was located that explicitly
assessed the impact of being Jewish. Rose (1983) had both Jewish and non-Jewish
children of survivor groups and Jewish and non-Jewish control groups in her study.
Okner and Flaherty (1988) were the only researchers comparing an American group of
children of survivors to an Israeli group.
As is outlined more fully in Chapter Four, the nature and duration of persecution
endured by survivors was highly dependent on the country in which they were
domiciled both before and during the war. It would therefore seem prudent to check for
differences between survivors of different countries of origin. The only study located
by the current author to have cross referenced results with country of origin was
Schleuderer (1990). Schluederer’s (1990) study examined children of survivors but he
cross-referenced their results with both their mother ‘s and their father’s country of
origin. No studies were located that examined this issue with the survivor generation
itself which is difficult to understand.
There is evidence for heterogeneity within the Holocaust survivor groups
assessed in the studies examined for this project. For example, the Holocaust survivor
groups assessed by Finer-Greenberg (1987) and Yehuda, Kahana, Southwick and Giller
Jr. (1994) both recorded standard deviations in depression scores twice as high as those
noted for their respective control groups.
© Janine Lurie-Beck 2007
While it is not possible to assess the
39
source/reason for this difference, it is possible to state that there was much more
individual variation in the Holocaust survivor groups compared to the control groups.
Perhaps if these groups were broken into demographic subgroups of some kind some
light could have been shed on the exact nature of these differences.
One possible explanation for Finer-Greenberg’s (1987) results is age
differences.
The age range of her Holocaust survivor sample was 17 years.
Specifically, participants ranged in age from 55 to 72 years. Those aged 55 would have
only been 7 years old when the war started in 1939, while those aged 72 would have
been 24 years old and already adults. Shmotkin and Lomranz’s (1998) sample of
Holocaust survivors had an even wider age range of 48 years, representing another lost
opportunity to ascertain the impact of age. Terno, Barak, Hadjez, Elizur and Szor’s
(1998) comparison of incidence levels of affective disorder among clinical groups of
survivors and controls is also called into question when one notes the age range of 38
years of their subjects.
While a lot of possible demographic moderating variables have been left
unexamined, some researchers have embarked on the assessment of differences between
survivors with and without a diagnosis of Posttraumatic Stress Disorder (for example
Trappler, Braunstein, Moskowitz, & Friedman, 2002; Yehuda et al., 1994; Yehuda,
Schmeidler, Giller et al., 1998). That any person diagnosed with PTSD would score
higher on measures of psychopathological symptoms such as depression, anxiety and
paranoia than persons without such a diagnosis seems to be a fairly logical assumption –
given the co-morbidity rates of these disorders. Research into differences between these
two groups does not add to our knowledge of what leads to differing levels of
adjustment within this population.
Many researchers have suggested various demographic factors which may have
influenced the degree to which a survivor (and their descendants) were able to adjust to
life after the Holocaust. These factors and the theories behind them will be explored
more thoroughly in Chapter Four.
2.6. – Summary and Conclusions
This chapter has provided a summary of the research and theories as to the impact of the
Holocaust on survivors and their descendants. As was discussed early in the chapter,
initial research focussed on the survivors themselves and then moved on to their
children as they started to seek treatment in the mid 1960s. The most commonly
referred to symptoms and effects of the Holocaust seen among survivors themselves are
depression, anxiety, paranoia, PTSD symptoms and difficulties with trust and intimacy.
© Janine Lurie-Beck 2007
40
Similar symptoms were seen among children of survivors in levels higher than the
general population.
If descendants of survivors also show elevated levels of psychological
symptoms then this begs the question of how did the effects of the Holocaust transfer to
subsequent generations. Chapter Three discusses numerous variables that have been
postulated as potential transmission modes. At the end of this chapter the preliminary
Model of the Differential Impact of Holocaust Trauma across Three Generations which
was presented in Section 2.4 will be expanded to include these proposed modes of
intergenerational trauma transmission.
© Janine Lurie-Beck 2007
41
Chapter Three – The Intergenerational Transmission of Holocaust Trauma
Surviving carries baggage that lasts for generations.
A child of survivors
Chapter Two cited many sources that suggest that the impact of the Holocaust can be
seen not only in those who directly endured the suffering of the period but in their
descendants as well. Related to this is the fact that the well-established finding that
higher rates of psychopathology are found among the children of people with
psychopathological symptoms or disorders has been replicated within the survivor
population as well (Keinan, Mikulincer, & Rybnicki, 1988; Major, 1990; Schwartz,
Dohrenwend, & Levav, 1994; Yehuda, Halligan, & Bierer, 2001). But it is not enough
to merely note that the greater prevalence of psychological symptoms noted amongst
survivors can be seen in their descendants.
It is necessary to investigate the
mechanisms that allow this to occur. Perhaps if these mechanisms can be identified
then it may be possible to nullify or minimise their effects and by so doing decrease the
symptom levels of the children.
This line of thought brings the issue of transmission of trauma across
generations, or intergenerational transmission of trauma to the fore. Intergenerational
transmission implies that children of trauma survivors can be exposed to or become
subject to “residues” of parental traumas without direct exposure to the trauma
(Weingarten, 2004). If it is accepted that descendants of survivors develop symptoms as
a result of their parents’ war time experiences, the next question is by what process or
processes does this occur?
Family systems theory applies to the case of symptomatology experienced by
descendants of Holocaust survivors quite readily. The family systems approach states
that as a part of an organised family system, an individual is never truly independent
and can only be understood within the context of that family system (P. Minuchin,
1985). A family systems approach to the study of psychopathology holds that an
individual’s symptoms are an expression of some form of dysfunction within the family
system which needs to be explored as much as the individual clients themselves (Bitter
& Corey, 1996). Certainly the nature of their survivor parents’ Holocaust experiences,
and the ways in which these impacted on their interactions with their children, are
highly relevant when considering the presenting symptomatology of children of
survivors. As Kellerman (2001d) argued “whatever the diagnosis, it is assumed that
children have absorbed some of the terrors of their parents and that any psychological
© Janine Lurie-Beck 2007
42
evaluation of the child must include some inquiry into the Holocaust history of the
parents as well as their particular style of child-rearing behaviour (p. 2).”
A number of hypotheses about survivor family environment have been posited
to explain why children of Holocaust survivors may have higher levels of symptoms
than the general population. These hypotheses fall into two camps. The first proposes a
direct form of transmission whereby the children of survivors develop symptoms via
modelling their parents’ maladaptive behaviours as might be espoused by theorists such
as Bandura (Kellerman, 2001a, 2001e; Major, 1990; Mazor & Tal, 1996; Schwartz et
al., 1994). The second suggests a more indirect route in which survivor’s parenting
behaviour is affected by their symptoms and the children subsequently develop
symptoms as a result of this “flawed” parenting (Felsen, 1998; Goldwasser, 1986;
Kellerman, 2001a; Schwartz et al., 1994; J.J. Sigal, 1973). In this second proposed
pathway interactions and communication within the family are seen as mediating
processes for the transmission of survivors’ psychological problems to their children
(Kellerman, 2001a). It is inevitable that survivors of any form of massive trauma such
as the Holocaust unwittingly incorporate their process of coping with that trauma into
their family life (Bistritz, 1988; Brom et al., 2001; Chaitin, 2002). Last (1989) notes that
this process was termed “toxic parental impact” within psychodynamic circles. The fact
that Holocaust survivors were seen to have difficulty in interpersonal relations after the
war has already been discussed in Chapter Two. It follows that these difficulties would
translate into their relationships with their post-war children and families. In other
words, the trauma does not only affect the individual survivor but also by affecting the
way the individual survivor interacts within the family unit it also affects the
functioning of the family unit, as would be espoused by family systems theory (P.
Minuchin, 1985). However, it should also be noted that, as Newman (1979) points out,
not all survivors developed problematic relationships with their children just as not all
survivors developed debilitating psychiatric problems in the post-war period.
Harkness (1993) described the family life cycle in a way that captures the reason
why the Holocaust can have such a long lasting and devastating impact on a family. He
said that “the family life cycle is characterised by the making and breaking of emotional
bonds”. Bar-On (1995) further elaborates by stating that families are unique systems
that can only be joined through birth or marriage and left through death or divorce.
Wardi (1994) adds that the “experience of loss and separation accompanies us
throughout our lives” and that “the ability to pass through life’s stages in a relatively
easy manner derives from the ability to experience separation in an optimal manner, as
© Janine Lurie-Beck 2007
43
conflict free as possible”. It is fair to say that survivors were not able to experience the
separation from their loved ones that perished during the Holocaust in a smooth and
conflict free way.
Wardi (1994) contends that as a result of this, developmental
separations and losses are especially difficult for Holocaust survivors to cope with.
Kellerman (2001b) further asserts that it is specifically in the area of making and
breaking emotional bonds that Holocaust survivor families often struggle.
In his
Person-Environment Approach to Traumatic Stress Reactions, Wilson (1989) cites
“intensification of developmental stages” as being part of life-course development in a
post-trauma environment.
From the outset survivors had difficulty establishing bonds after the Holocaust
and they subsequently ran into difficulty when those bonds had to be severed. This
corresponds to difficulty in establishing parent-child attachment and more generally to
cohesion within the family and the children’s separation-individuation phase
corresponds to the breaking of bonds. Parent child attachment, family cohesion and
separation-individuation have been mentioned many times in the literature as
particularly difficult for survivors and their families. Additionally, communication
within the survivor family, particularly in relation to the Holocaust experiences of the
survivor, has been presented often as another mediating influence on the psychological
adjustment of children of survivors.
This chapter considers the four factors of parent-child attachment, family
cohesion, separation-individuation and communication, in turn in terms of their nature
within survivor families and their impact on the well-being of the children in these
families. In so doing this chapter serves to incorporate a broad family systems approach
to the assessment of transmission of Holocaust trauma incorporating attachment theory
and the 3-D Circumplex Model of Family Functioning (cohesion and communication).
Finally, the preliminary Model of the Differential Impact of Holocaust Trauma across
Three Generations presented in Chapter Two, Section 2.4, will be amended to
incorporate these proposed intergenerational transmission variables.
3.1. – Parent-Child Attachment
Chapter Two highlighted the fact that many survivors experienced problems in relating
to others after the war (for example Freyberg, 1980). Sigal, Silver, Rakoff and Ellin
(1973) explain how this deficit among survivors lead to difficulties in forming healthy
relationships with their children. There has been an increasing emphasis in the literature
on the impact of traumas such as the Holocaust on attachments (Wiseman et al., 2002)
© Janine Lurie-Beck 2007
44
and the use of attachment as a useful framework in the explanation and treatment of
traumatic impact (for example M. F. Solomon & Siegel, 2003).
The study of attachment focuses on the infant as an individual “in formation”
within the family system (P. Minuchin, 1985). A very basic definition of attachment is
that it is the emotional bond that forms between an infant and a parent/adult that leads
the child to seek out that adult in times of distress, be highly receptive to being cared for
by them and to display anxiety if separated from them (Reber, 1985). The child can
either develop a secure or insecure attachment to their parent or caregiver depending on
the nature of the adults’ responses to the child.
If the parent or caregiver responds
promptly and consistently to the child’s needs then a secure attachment will form. If the
parent or caregiver is slow to respond, inconsistent in their response or non-responsive
then an insecure attachment will form.
Inconsistency of response will lead to an
ambivalent attachment while non-responsiveness will lead to avoidant attachment (the
work of Mary Ainsworth as summarised by Hazan & Shaver, 1987). A child’s
attachment to their parent/s is very important as it teaches the child emotional
regulation. This role makes it clear how attachments in formative years can be linked to
emotional problems in later life (Sroufe, Duggal, Weinfeld, & Carlson, 2000).
The role that the parent-child relationship may play in the mental health of the
children is also a basic tenet of Kohut’s Self-Psychology theory. Kohut argued that the
failure of parents to demonstrate sufficient empathy towards a child is a key element in
the development of psychopathology (the work of Heinz Kohut as summarised by
Baker & Baker, 1987).
A lack of responsiveness or inconsistent responsiveness,
mentioned in the context of attachment theory above, also relates to the selfpsychological concept of parental mirroring. Adequate/appropriate responsiveness or
mirroring is an important resource for the development and maintenance of self-esteem
in the child, and when it is inadequate there is a much higher likelihood of the child
having psychological problems in adulthood (the work of Heinz Kohut as summarised
by Baker & Baker, 1987).
3.1.1. – Survivor Parents’ Insecure Attachment to their Children
The psychological symptoms experienced by survivors in the post-war period often had
negative implications for the way in which they could relate to their children. In
particular problems with unresolved mourning and anxiety have been linked to
problematic parenting and attachments. Unresolved mourning in mothers has been
linked with disorganised infant-mother attachments (van Ijzendoorn, 1995). Parents
© Janine Lurie-Beck 2007
45
pre-occupied with their grief over lost relatives were often emotionally unavailable to
their children (Barocas & Barocas, 1980).
Depressed and anxious parents have also been seen to be withdrawn and lacking
in warmth towards their children which has implications for the parent-child
relationship (Kendler, Sham, & MacLean, 1997). An anxious parent is limited in their
ability to inspire trust in their children (Lipkowitz, 1973).
As noted in Chapter Two, anxiety is commonly found among survivors and
survivors have often been described as over-anxious parents (Axelrod et al., 1980; Halik
et al., 1990; Kellerman, 2001b; Newman, 1979; Rustin, 1988). Davidson (1980a)
describes how this over-anxiety was played out in the mother-child relationship from
the very beginning with survivor mothers overfeeding their babies for fear they would
starve, constantly checking on babies’ breathing throughout the night with extreme over
reactions to any minor sign of ill health. Sonnenberg (1974) explains that survivors
were driven by a great desire to protect their children from having to endure any
suffering and Halik et al. (1990) add that their fear that harm would come to their
children was often inappropriately extreme.
Children of survivors often note that their parents were hyper-vigilant and
protective (Freyberg, 1980). It is not surprising that such behaviour was found to
inspire fear, mistrust, wariness and suspicion in children of survivors (Rowland-Klein &
Dunlop, 1997). While this level of anxiety may have been appropriate for conditions
during the Holocaust it is considered inappropriately extreme for peacetime conditions.
For example, Barocas and Barocas (1980) write that survivors often gave the impression
of extreme and imminent danger though its source may have seemed fairly vague
(Kellerman, 2001a).
They made general statements such as “be careful” and “don’t
trust anybody” (Kellerman, 2001e). Kellerman (2001a) suggests that some survivors
continued with their hyper-vigilant survival strategies which may well have been key to
their survival in a concentration camp but are viewed as paranoid in a peacetime
environment.
Survivors often have a retrospectively insecure view of their relationship to their
own parents. This is not because the nature of the parent-child attachment was insecure
but because the intervening circumstances of the Holocaust altered their assessment.
Research has found that attachment can change over time and can be considered
unstable (Baldwin & Fehr, 1995). Therefore it can be argued that a traumatic event and
how a parent is seen to deal with that event can alter the nature of the attachment that
the child has with their parent. For example, child survivors’ images of protecting and
© Janine Lurie-Beck 2007
46
loving parents were undermined when they witnessed their parents’ inability to protect
themselves or their children from Nazi atrocities (Chaitin & Bar-On, 2002; Kestenberg
& Kestenberg, 1980). Thus the traumas of the Holocaust induce an insecure attachment
to survivors’ parents which may not have existed prior to these events (Auerhahn &
Lamb, 1998).
Some survivors were able to maintain their image of their pre-war
relationship/attachment to their parents as a model for their later child-rearing.
Survivors who were able to hold on to these memories were more likely to be able to
foster secure attachment in their children as they had a model of security they could
convey to their children (Chaitin & Bar-On, 2002). Survivors who did not have such an
internalised model often had difficulties in this regard (Freyberg, 1980).
However, van Ijzendoorn, Bakermans-Kranenburg, and Sagi-Schwartz (2003)
disagree that survivors retrospectively altered their view of their attachment to their
parents. They argue that while they certainly saw evidence of their parents’ inability to
protect them, survivors were able to recognise that the fault for this lay with the Nazi
war machine and not with any parental weakness. They argue that survivors attachment
to their parents and basic trust should not have been undermined by their Holocaust
experiences (van Ijzendoorn et al., 2003).
Whatever the argued origin, be it unresolved mourning, over-anxiety or
retrospective alterations of attachments based on Holocaust experiences, it has been
argued by many that the parent-child attachment between survivors and their children is
more often insecure than secure.
3.1.2 – Negative Effects of Insecure Attachment of Children to Parents
Insecure modes of attachment between parents and children have been linked to
increased symptomatology in the children such as depression, anxiety, paranoia and
adult attachment insecurities (Crowell & Treboux, 1995; Hesse, Main, Yost Abrams, &
Rifkin, 2003). In a recent study, Costa and Weems (2005) found that attachment to
mother mediated the relationship between child and maternal anxiety. In other words,
the apparent relationship between maternal and child anxiety was reflecting the
relationship between maternal anxiety and her attachment to her child and the child’s
attachment to their mother and their level of anxiety. This mediation relationship was
also found by Han (2005) with a sample of children of South East Asian refugees who
had been traumatised by war experiences. She suggested that because the refugee
parents had a reduced capacity to develop secure attachment with their children (in
much the same way that the current author and the other authors have argued is the case
© Janine Lurie-Beck 2007
47
for Holocaust survivors) that therefore parent-child attachment can be considered a
pathway for trauma transmission (Han, 2005).
Having an anxious or depressed parent has been linked to children’s insecure
attachment to those parents. For example, Radke-Yarrow (1991) found the incidence of
secure attachment to be lower among children of depressed mothers than of mothers
with no diagnosis (53% versus 62%).
Shoshan (1989) notes that some children of
survivors felt their parents minimised the importance of any problems they were having
for example at school and so felt that they could not turn to their parents for comfort and
support in these instances. Such behaviours are linked to insecure attachment modes
(van Ijzendoorn, 1995). Survivor parents have often been described in this way (Chaitin
& Bar-On, 2002; Freyberg, 1980; Lipkowitz, 1973).
Bar-On et al. (1998) argue for the application of the categorisation of attachment
between survivor parents and their children as the insecure-ambivalent attachment
orientation. Main (1990) stated that this form of attachment is the predictable response
of a child to the inconsistency of emotional responsiveness of the survivor parent.
Survivors have been described as clinging tightly to their loved ones for fear of losing
them on the one hand and at the same time seeming to distance themselves so that they
would not be so hurt if they did lose them (Brom et al., 2001). The children in this
situation are then conflicted by their desire for emotional proximity which they often
felt denied. However they also strove for autonomy in reaction to the stifling nature of
their parent’s over-anxiety and protectiveness. This description melds neatly with the
insecure-ambivalent attachment style. Ryn (1990) notes that relationships between
survivors and their children have been described as ambivalent in Polish research. The
attachment between survivors and their children has been described as symbiotic by
many writers (Barocas & Barocas, 1980; L. Berger, 1988; Halik et al., 1990; Kellerman,
2001c; Rowland-Klein & Dunlop, 1997; Rustin, 1988).
The contrasting model of Fraley, Davis and Shaver (1998) which explains the
rise of a dismissing/avoidant attachment style can also be applied in some survivor
families. In their model, Fraley et al. (1998) note how a rejecting caregiver who is
“unavailable” to their children leads their children to feel they cannot rely on others and
not to become too close or dependent on others. This model in particular would apply
in survivor families where survivors remained emotionally distant to their children
because of a fear of pain caused by further separations. The theory being that if they
tried to remain detached they would not be hurt so much when the inevitable separation
occurred.
© Janine Lurie-Beck 2007
48
These problematic parent-child relationships can then be linked to symptoms in
the children such as depression (J.J. Sigal et al., 1973). Feeney and Noller (1996) state
that quality of parent-child attachment has been found to relate to emotional well-being
of children with lower anxiety and depression associated with secure attachment and
higher anxiety and depression associated with insecure forms of attachment It has been
argued that the children of over-anxious parents have higher incidence levels of
depression, paranoid ideation, phobias and psychoses (Rustin, 1988; Trossman, 1968).
The findings of a study by Vivona (2000) further attest to this relationship with
adolescents reporting insecure attachments to their parents evidencing higher levels of
depression and anxiety. Sroufe, Duggal, Weinfeld and Carlson (2000) assert in their
review that the relationship between insecure parent-child attachment and emotional
problems among the grown up children is firmly established. However they temper this
by saying that this relationship is not deterministic but rather probabilistic. Emotional
or psychopathological problems are more likely when insecure-parent child attachment
is present but are not a certainty.
The insecure attachment that children of survivors developed towards their
parents is also related to attachment problems in later life (Crowell & Treboux, 1995).
Some theorists argue that adult attachment is a natural progression or evolution of
attachments to parents in childhood and that the only thing that changes is the object of
that attachment, that is from the parents to the romantic partner (Levy, Blatt, & Shaver,
1998; R. S. Weiss, 1991). Bowlby (1988) contended that the attachment that a child
forms with its mother or other primary care giver acts as a model for subsequent
relationships with romantic attachment figures. Davidson (1980a) argues that children
of survivors often have difficulties establishing intimate relationships because of lack of
security and warmth they felt as a child. This may be because they were lacking a
secure attachment model (E. Cohen et al., 2002).
Numerous studies have found
evidence of a relationship between insecure parent-child attachment and insecure
romantic attachment among grown children (Collins & Read, 1990; Levy et al., 1998;
Shaver, Hazan, & Bradshaw, 1988). Sagi-Schwartz et al. (2003), Woolrich (2005) and
Berger (2003) obtained further support for this relationship within the survivor
population.
Not only are children of survivors’ adult attachments potentially affected, but
also their subsequent attachments to their own children. Thus attachment becomes a
self-perpetuating transmission mode (Chaitin & Bar-On, 2002; van Ijzendoorn, 1995).
In a meta-analysis of parent-child attachment van Ijzendoorn (1995) found that in 75%
© Janine Lurie-Beck 2007
49
of cases, the nature of parent-child attachment was predicted by the parent’s view of
their own childhood attachment. If they viewed their own childhood attachment as
insecure then their attachment with their children is also likely to be insecure. Since
insecure attachment has already been linked to the increased likelihood of psychological
problems such as anxiety and depression Sigal and Rakoff (1971) note that there is
every possibility that children generations removed from the Holocaust will still be
suffering from its psychological aftermath.
3.2. – Family Cohesion
Family cohesion can be defined as the degree of emotional bonding or closeness
between family members (Bray, 1995; Olson, 1993). In other words, how close or
distant family members feel to each other (Bray, 1995). According to Olson’s (1993)
Circumplex model of marital and family systems, the four levels of cohesion are
enmeshed (very high), connected (moderate to high), separated (low to moderate) and
disengaged (very low). The two extremes of this spectrum are considered maladaptive
while the two central dimensions are considered normal (Olson, 1993).
3.2.1. – Extreme Levels of Cohesion in Survivor Families
Family cohesion has been heavily implicated in theories of trauma transmission
(Baranowsky et al., 1998). For example, Leydic Harkness (1993) notes that families of
Vietnam veterans have been categorised at the extreme (maladaptive) ends of the
cohesion dimension, that is enmeshed or disengaged.
The extreme ends of this
dimension have been linked to psychological problems such as depression in the
children of these families (Franklin & Streeter, 1993).
Holocaust survivor families are commonly described as being enmeshed or
disengaged, though enmeshed is the more common description (for example Almagor &
Leon, 1989; Felsen, 1998; Halik et al., 1990; Kellerman, 2001b, 2001c; Newman, 1979;
Perel & Saul, 1989; J.J. Sigal & Weinfeld, 1989). Kellerman (1999) notes that survivor
families are often infused with over-dependency between family members or
exaggerated independence. In a later article he notes that while survivor families differ
the more pathological ones can be considered enmeshed (Kellerman, 2001e).
According to Olson (1993) disengaged families are characterised by a lack of
emotional closeness, with members generally unable to turn to each other for support.
Family members see themselves as quite separate and have a high degree of
independence with little attachment to the family. The emotional closeness lacking in
the disengaged family system is omnipresent in the enmeshed family. Dependency
between family members is very high and family members have very little identity as a
© Janine Lurie-Beck 2007
50
separate person. Minuchin (1974) contended that the parents in enmeshed families are
often overcontrolling and intrusive when relating to their children.
Holocaust survivor parents have often been characterised as over controlling and
specifically over protective of their offspring (Axelrod et al., 1980; Bar-On et al., 1998;
Chodoff, 1997; B. B. Cohen, 1991; Felsen, 1998; Fogelman & Savran, 1979; Halik et
al., 1990; Hass, 1990; Kellerman, 2001b; Kuperstein, 1981; Newman, 1979; Oliner,
1990; Rowland-Klein & Dunlop, 1997). In fact, in his discussions with children of
survivors, Hass (1990) found that over-protectiveness was the most common trait
mentioned when asked how they believed their parents’ Holocaust experiences had
affected the way they raised them. Hogman (1998) asserts that over-protectiveness
towards children was also noted among survivors of the Armenian genocide.
It is worthy of mention that while enmeshment is generally viewed as a negative
influence, Fogelman and Savran (1979) note that children of survivors do derive some
positives out of it. They note that while the children of survivors they have come into
contact with often express anger and frustration over their parents’ over-protectiveness
they also see it as evidence of their parents’ deep love for them (Fogelman & Savran,
1979). Because of the extent of their losses in the Holocaust, survivors value their
family members to a great extent and this cannot be but obvious to their children
(Fogelman & Savran, 1979).
3.2.2. – Extreme Family Cohesion as it Relates to the Psychological Health of Children
Over-protection on the part of parents has been linked to depression among their
children in the general population (Feeney & Noller, 1996). Rustin (1988) posits that
over-protection on the part of survivor parents has been linked to depression and
paranoid ideation in children of survivors.
Sachs (1988 (in Hebrew), cited in Z.
Solomon, 1998) found that extreme cohesion (either enmeshment or disengagement)
was related to higher levels of anxiety and depression among children of survivors.
As Almagor and Leon (1989) and Felsen (1998) note, the lack of boundaries
both between the parents and intergenerationally between the parents and the children
inevitably lead to difficulties in the establishment of self-identity in the children of these
families. Ackerman (1956, cited in Almagor & Leon, 1989) suggested that this blurring
of boundaries leads to “interlocking of family pathology as well as its generational
transmission.”
While high and low cohesion have been linked to psychopathology symptoms in
the children of affected family systems it is the decreased level of independence
associated with high cohesion that is more often discussed in relation to children of
© Janine Lurie-Beck 2007
51
Holocaust survivors. When cohesion is too high, or a family is considered enmeshed,
family members have minimal independence (Olson, 1993). This lack of independence
can cause problems when children enter the separation-individuation phase during
adolescence (Felsen, 1998). Specifically the children have problems establishing
autonomy and independence (Felsen, 1998; J.J. Sigal & Weinfeld, 1987). These
difficulties are discussed in the next section.
3.3. – Separation-Individuation
According to Mazor and Tal (1996), separation-individuation refers to the development
of an autonomous, responsible adult identity separate from the family of origin able to
take on adult responsibilities. This occurs gradually as the balance between autonomy
and independence and dependence on parents is adjusted in age-appropriate stages
(Mazor & Tal, 1996). The end result is commonly physically demonstrated by the child
moving out of the family home in adolescence or young adulthood (Auerhahn & Lamb,
1998).
Emotional fusion is the polar opposite of individuation and is characterised by
unresolved attachment to the family of origin (Bray, 1995). Bray (1995) explains that
children at this end of the spectrum feel undue responsibility for others (that is, their
parents) and may also avoid taking responsibility for themselves by maintaining such
strong ties to their parents. Such fusion is linked to higher stress levels, poorer health,
and higher incidence of psychological problems (Bray, 1995; Brom et al., 2001).
3.3.1. – Separation-individuation problems noted among children of survivors
Many researchers have noted that children of survivors have extreme difficulty
establishing an independent and autonomous self (Chaitin & Bar-On, 2002; Gay et al.,
1974; Newman, 1979; Perel & Saul, 1989). In fact, Perel and Saul (1989) state that
stages of family development that involve assertion of autonomy and independence,
especially when children reach adolescence, leave home and marry, are particular crisis
points for the survivor family. However Wardi (1994) opined that survivor mothers are
incapable of encouraging independence even in their infant children. With the blurring
of boundaries outlined in previous sections in the form of symbiotic attachments and
family enmeshment it is no wonder that individuation presents difficulties for survivors
and their children and that fusion is prevalent (Bar-On et al., 1998; Barocas & Barocas,
1980; Brom et al., 2001; Felsen, 1998; Fogelman & Savran, 1979; Freyberg, 1980;
Goldwasser, 1986; Halik et al., 1990; Hass, 1990; Jucovy, 1992; Kellerman, 2001b,
2001e; Kuperstein, 1981; Mazor & Tal, 1996; Perel & Saul, 1989; Rosenbloom, 1988;
Rowland-Klein & Dunlop, 1997; Shoshan, 1989; Steinberg, 1989; Wardi, 1994).
© Janine Lurie-Beck 2007
52
The root cause of separation-individuation problems in the children of survivors
is their parents’ intense reaction at the thought of separation from them. This is thought
to result from the losses experienced by survivors during the Holocaust, particularly of
their parents (Halik et al., 1990; Kellerman, 2001e; Mazor & Tal, 1996). It is not the
fact that they were separated from their parents but that the separation was so abrupt and
violent that leads to difficulties (Last, 1989; Perel & Saul, 1989). Children were often
dragged away from their parents kicking and screaming or witnessed their parents being
killed (Chaitin & Bar-On, 2002). Having been separated from their parents in such an
unnatural way, many survivors are lacking a parental model for dealing with the normal
process of individuation (S. Davidson, 1980a).
They interpret separations as loss
(Freyberg, 1980) since in their experiences separations between parents and children
were permanent. Also having lost so many relatives as well as the collective loss of
communities, many survivors feel incapable of dealing with any further loss (Kahana et
al., 1989; Rowland-Klein & Dunlop, 1997).
Hass (1990) states that some survivors explicitly told their children that they
could not endure another separation, even the normal developmental separation of
young adults from their family of origin. Some go so far as to block their children’s
attempts at establishing relationships with people outside the family unit, to move out of
the family home or any other attempts at individuation (Barocas & Barocas, 1980; S.
Davidson, 1980a; Freyberg, 1980; Hass, 1990). When this occurred the children often
developed an ambivalent relationship with their parents due to the resentment they felt
from their parent’s clinginess and their interpretation of their attempts at individuation
as acts of abandonment and betrayal (Bistritz, 1988).
Children of survivors develop “separation guilt” as a result of their desire to
differentiate themselves from their parents (Grubrich-Simitis, 1981; Kuperstein, 1981;
Rowland-Klein & Dunlop, 1997). Children of survivors are torn between wanting to
establish relationships outside the family and establish a degree of autonomy and
independence yet at the same time wanting to maintain the comfortable dependency on
their family of origin (Russell, 1982). They are aware of how much their parents have
already suffered and lost and so wish to protect them from further separations
(Freyberg, 1980; Mazor & Tal, 1996; Rowland-Klein & Dunlop, 1997). Many children
of survivors therefore remain at home long after many of their cohort have moved out
on their own (Fogelman & Savran, 1979) thereby maintaining their symbiotic,
enmeshed relationship with their family of origin to the detriment of their own
autonomy (Rowland-Klein & Dunlop, 1997). Thus just as it was argued earlier in this
© Janine Lurie-Beck 2007
53
chapter that survivors are overly protective of their children, so too are children of
survivors of their parents (Kellerman, 2001e).
3.3.2. – Relationship between Separation-individuation Problems and Negative
Psychological Outcomes
It has been stated that the difficulties experienced in this stage are the main cause of
symptoms in the children of survivor population (Grubrich-Simitis, 1981). Separationindividuation is a major goal of the adolescent period (Kuperstein, 1981) and it was at
this life stage that children of survivors started to seek psychological help in greater
numbers (S. Davidson, 1980a; Kuperstein, 1981; Russell, 1982).
Problems with separation-individuation and the aforementioned guilt that
children experience have been associated with a number of psychopathological
symptoms such as depression, anxiety and paranoia (Barocas & Barocas, 1980; Bistritz,
1988; S. Davidson, 1980a; Okner & Flaherty, 1988; Rustin, 1988; Steinberg, 1989).
Kenny and Donaldson (1991) report statistically significant positive relationships
between children’s symptom levels and parental over-involvement and parental fear of
separation from their children, while a negative relationship was noted between
children’s symptoms and the degree to which parents encouraged their autonomy.
Mazor and Tal (1996) note that the capacity to develop intimacy with a romantic
partner is also related to the degree to which an individual is differentiated from their
family of origin. Along with Steinberg (1989) and Barocas and Barocas (1980), they
note that when attachments to family of origin have not been sufficiently processed and
resolved via the separation-individuation process, this style of symbiotic attachment
may then be played out in future attachment relationships both romantic and parental.
3.4. – Communication
Survivors’ ability to talk about their Holocaust experiences varies greatly. Of specific
interest for this project is the way in which survivors chose to communicate their
Holocaust experiences to their children. Variations in the communication of Holocaust
experiences can have implications for the psychological well-being of survivors’
children (Axelrod et al., 1980; Baranowsky et al., 1998; Bistritz, 1988; Okner &
Flaherty, 1988; Wiseman et al., 2002). The influence of the style of communication
used by other trauma victims such as Vietnam veterans and ex-prisoners-of-war has also
been found to impact on their descendants. Kellerman (2001e) states that parental
communication style has been implicated as “a crucial determinant in the adaptation of
families beset by catastrophe.” Bray (1995) contends that “communication deviance”
has been linked with disorders such as schizophrenia.
© Janine Lurie-Beck 2007
54
Open communication in general is linked to the more functional family systems
while dysfunctional families commonly have communication problems (Bray, 1995).
Family communication patterns can be seen as symptomatic of a family’s level of
cohesion. Bray (1995) posits that families with high cohesion, or enmeshment, display
excessive emotional responsiveness when relating to family members, while families
with low cohesion, or disengaged families, may have problems with a lack of
communication.
For example, Rodick, Henggler and Hanson (1986) found that
interaction between mothers and adolescents characterised as balanced was associated
with statistically significantly higher rates of supportive and open communication
expressed with positive affect than those located at the extremes. In fact Olsen (1993),
includes family communication as a facilitating dimension in his Circumplex Model of
Family Systems, which aids in movement along the two dimensions of family cohesion
and adaptability.
Communication between survivors and their children about the Holocaust has
varied between complete and (too) detailed accounts to absolute silence (Baranowsky et
al., 1998; Chaitin & Bar-On, 2002; Danieli, 1988; Fogelman & Savran, 1979; Gay et al.,
1974; Gertz, 1986; Jucovy, 1992; Jurkowitz, 1996; Kellerman, 2001a, 2001c, 2001e;
Obermeyer, 1988; Okner & Flaherty, 1988; Perel & Saul, 1989; Rowland-Klein &
Dunlop, 1997). Communication in one of these extreme forms has been linked to the
least adjusted of the children of survivor population (S. Davidson, 1980a; Jucovy, 1992;
Kellerman, 2001a; Rieck, 1994).
While open communication is seen as favourable, it has been argued that
tempering of accounts is required so as not to traumatise the listener with horrific detail
when discussing the Holocaust. Obsessive talk/preoccupation about the Holocaust does
not take into account the listener’s ability to absorb and can be traumatic for the listener
(Bar-On, 1995; Baranowsky et al., 1998; Kellerman, 2001c; Perel & Saul, 1989).
Schwarz (1986) found a curvilinear relationship between children’s depression and the
level of parental communication about their Holocaust experiences. Children forced to
continuously listen to detailed accounts of their parent’s suffering often become
depressed or develop a strong sense of guilt over not having suffered like their parents
(S. Davidson, 1980a; Kuperstein, 1981; Trossman, 1968). Okner and Flaherty (1988)
and Major (1990) note that the age at which children of survivors are told about their
parents’ experiences is also of importance. This point ties into the aforementioned idea
that the listener’s ability to absorb what is being told is important. If children are told at
a very young age when they are not ready to optimally process the information it can
© Janine Lurie-Beck 2007
55
have a negative impact on their psychological health (Major, 1990; Okner & Flaherty,
1988).
Major (1990) notes that among children of Norwegian survivors who
incessantly spoke about their Holocaust experiences when the children were very young
depression and nightmares were common. She suggests this is because the children
were unable to keep a healthy distance from the experiences they learnt about from their
fathers’ Holocaust stories (Major, 1990).
The conspiracy of silence that was quite pervasive for around a decade (as
discussed in Chapter Two) meant that a lot of survivors had no one to talk to about their
experiences (Danieli, 1982; Kellerman, 2001e; Wiseman et al., 2002). Therefore the
survivors’ children became the only sounding boards available for a lot of survivors
(Gordon, 1990). This factor contributed to survivors seemingly obsessive re-telling of
Holocaust stories (Gordon, 1990). For many survivors the telling or witnessing of their
Holocaust experiences was a necessary form of catharsis (Danieli, 1982) as it presented
a way of releasing pent up grief (Okner & Flaherty, 1988). Because survivors often felt
they could not or did not want to seek any form of psychological help the children were
counted on for emotional support as well (Fogelman, 1998). For all of these reasons the
children in these circumstances were understandably affected.
Silence on the topic can also be harmful (S. Davidson, 1980a). Many survivors
remained silent about their experiences in the belief that their children would be
traumatised if they told them about it (Baranowsky et al., 1998; Danieli, 1988; Jucovy,
1992; D. Weiss, 1988; Wiseman et al., 2002). In fact Finkelstein and Levy (2006) state
that 22% of their survivor sample (n = 50) cited “fear of audience harm” as a reason for
their reluctance to talk about their Holocaust experiences. However, when the children
are not told anything about their parents’ experiences they often play out fantasies in
their head of what they think might have happened (S. Davidson, 1980a; Major, 1990).
These fantasies may be more horrific than the truth (Baranowsky et al., 1998; L. Berger,
1988; S. Davidson, 1980a; Jucovy, 1992; Steinberg, 1989). Therefore their parents’
silence indirectly had the opposite effect to that desired, namely to cause psychological
distress (Kellerman, 2001a; van Ijzendoorn et al., 2003) such as increased depression
(Major, 1990; Trossman, 1968). Goldwasser (1986) notes that an important part of
therapy for children of survivors is to learn to differentiate between the reality of their
parents’ experiences and the fantasies they created.
It seems that “the efforts by
survivor parents to protect their children from knowing the grotesque experiences they
had witnessed and endured ultimately failed in light of the child’s urge to know (Krell et
al., 2004, p 507).” Indeed a study by Wiseman et al. (2002) found that children of
© Janine Lurie-Beck 2007
56
survivors who had largely remained silent about their experiences were more anxious
than children of survivors who had openly discussed their experiences. Similarly,
Goodman (1978) found that a clinical group of children of survivors reported less
communication about the Holocaust by their parents than a group of children of
survivors classified as non-clinical.
Despite their parents’ silence on the topic of the Holocaust most children could
sense their parents’ suffering and were left confused and inexplicably guilt ridden about
it (Danieli, 1988). For example, from a very early age, perhaps from birth, children of
survivors can be ultra sensitive to non-verbal signals such as facial expressions which
belie the suffering hidden under the surface (S. Davidson, 1980a; Fogelman, 1989;
Kellerman, 2001e; Shoshan, 1989; Steinberg, 1989). Fogelman (1989) contends that it
is via these intense non-verbal messages teamed with small bits of information gathered
from various sources that a lot of children find out or piece together their parent’s
Holocaust story rather than via direct story telling. The following quote illustrates the
point:
Because silence transmits its own messages, it is impossible not to
communicate. Meanings are constructed. Snippets of text and
fragments of allusion are calibrated against context and sense is
haltingly induced. I grew up apprenticed in the skills of inference and
versed in the language of the oblique. I became literate in the
grammar of silence.
Wajnryb (2001, p. xi), a child of survivors
Children of survivors in this situation often don’t press their parents for details
as they sense that it would cause pain for their parents to discuss their experiences
(Wiseman et al., 2002). Bar-On et al. (1995) and Wiseman et al. (2002) refer to this
situation as a “double wall” in which parents do not tell and children do not ask.
Children are also often conflicted about whether they want to know the details of their
parents’ Holocaust experiences (Fogelman & Savran, 1979). Fogelman and Savran
(1979) reported that when this issue came up in children of survivor group sessions
some intimated that they wanted to know how their parents survived but at the same
time did not know how they would cope with the knowledge once they got it. Or as
Wajnryb (2001, p. xii) put it: “there has been a struggle of competing interests: the
yearning to understand has vied uncomfortably with the urge not to know.” Janice
Friebaum’s poetry also portrays similar themes.
© Janine Lurie-Beck 2007
57
You were mute to free our spirits
You kept secrets to spare us grief
Yet we were tethered to your pain
and mourned for your losses.
How could you know your eyes spoke
volumes
And your silence painted pictures?
In perfect step we moved:
Not-too-close, not-too-much,
Not-so-soon, please don’t push.
If you think we didn’t care
We thought you didn’t either.
If you thought we didn’t need your burdens
We thought we wouldn’t trouble you with
ours.
You thought, we thought,
You assumed, and so did we.
All to keep us safe, all to keep you unhurt.
Excerpts from Silent Conversation by Janice
Friebaum © (a child of survivors)
Lang (1995) notes that survivor parents who tell little of their Holocaust
experiences are often also relatively silent on their pre-Holocaust life. Therefore the
children of these survivors not only live in a state of confusion over their parents’
Holocaust experiences they also have an increased sense of isolation from the lack of
family history available to them. Weiss (1988) argues that many children of survivors
interpreted their parents’ reluctance to discuss their experiences as a lack of closeness
and warmth which understandably impacted upon the quality of the parent-child
attachment.
It is the children of survivors who were able to be balanced in their
communication about the Holocaust (that is open but not in excessive detail until the
listener is ready to cope) that evidence the least psychological impact (S. Davidson,
1980a; Perel & Saul, 1989). Goldwasser (1986) notes how many authors in the field
stress that Holocaust experiences should be communicated to children in “an
appropriate and non-frightening way”.
It has been suggested that the degree to which a survivor is able to discuss their
experiences with their children is related to the survivor’s progress in working through
or coming to terms with their experiences themselves (S. Davidson, 1980a; Russell,
1982). A survivor who has largely processed and reconciled their experiences is in a
better position to discuss the Holocaust in a healthy way with their children (S.
Davidson, 1980a). For these reasons Rowland-Klein and Dunlop (1997) suggest that
survivors feel more capable of telling their stories to their children when they have
grown up and with the added benefit of the passage of time. Survivors are then better
able to tell their stories and the children are mature enough to cope with them. It is also
worthy of note that many survivors who constantly verbalised their Holocaust
experiences were responding to a “need to talk” that Janoff-Bulman (1992) likened to
© Janine Lurie-Beck 2007
58
intrusive recall (akin to the PTSD symptom of intrusion).
Janoff-Bulman (1992)
suggests that persistent discussion of the Holocaust most likely points to incomplete
processing of the traumas by the survivor.
The ability to disclose details of experiences and be expressive has also been
linked to psychological well-being (Finkelstein & Levy, 2006; Gordon, 1990; Russell,
1982; Tedeschi, 1999) as well as the development of a feeling of security and trust
(Axelrod et al., 1980). Gordon (1990) postulated that a survivor’s ability to articulate
their Holocaust experiences is the mediating variable between the actual trauma and
their psychological well-being.
Furthermore the ability to clearly articulate one’s
experiences becomes a mode of intergenerational transmission for interpersonal
difficulties with such factors as intimacy.
Gordon (1990) argued that children of
survivors use their parents’ method of expression as a model for their own and if their
parents have difficulty articulating or acknowledging their experiences so will their
children. She goes on to state that this ability is related to the development of intimacy
in relationships and it therefore is reasonable to expect that children of survivors will
therefore have difficulties in this regard (Gordon, 1990).
The fact that survivors sometimes have trouble clearly articulating their
Holocaust experiences leads them to ambiguous communication about this period to
their children (Kellerman, 2001e). Bray (1995) defines healthy communication as being
clear and direct and so this form of communication, which he would categorise as
indirect and masked, is considered dysfunctional Rowland-Klein and Dunlop (1997)
note that communication about the Holocaust can therefore be subtle and somewhat
coded. Gordon (1990) gives a good example of this and the negative impact it can have.
A Holocaust survivor may react negatively to the colour yellow because of its
association with the Star of David armband that Jews were forced to wear during the
war. If the survivor is able to clearly explain this connection to the child they learn that
it is this link to the armband that makes yellow distasteful to the survivor and are able to
clearly understand this connection. If the survivor is unable to clearly articulate this
connection and responds in a fearful manner to a vast number of things that are yellow
or to various sounds and smells the child may take on this sense of a more globalised,
objectless fear or paranoia without understanding its source. Gordon (1990) argues that
it is not the automatic fear reaction witnessed by the child that is the important
transmitting factor but the survivors’ ability to explain it that is the key. When the
explanation is present, the necessary condition of healthy communication, that is shared
© Janine Lurie-Beck 2007
59
meaning, is also present (Bray, 1995). Lack of shared meanings is a feature of
dysfunctional communication (Bray, 1995).
However it is not simply a matter of being open or the extent to which a survivor
is open about their experiences that impacts on the psychological well-being of the
children of survivor population.
The way in which survivors communicate their
experiences can be very important. Goldwasser (1986) explains how the parent’s ability
to openly describe their experiences without being excessively emotive or alternatively
being too deadpan (blunting of affect) is of vital importance. Jucovy (1992) states it is
“the way in which communication between the generations takes place rather than the
concrete content which is being imparted” that is the influential factor.
Bray (1995) refers to the concept of emotional expression which applies here.
Bray (1995) notes that affect or emotion that is tied to verbal communications via things
such as tone of voice can impact on the way the message conveyed is interpreted. He
notes that strong negative emotions, termed expressed emotion when teamed with
verbal communication, have been seen in families of schizophrenic, depressed and bipolar patients and have been associated with increased depression and anxiety. He also
notes that strong negative emotional statements outweigh positive emotional statements
in terms of their impact on family life. Guilt-inducing communication has frequently
been noted within Holocaust survivor families (for example Kellerman, 2001e;
Lichtman, 1983; Porter, 1981). Porter (1981) notes that depression often results in
children of survivors as a result of this, while Keller (1988) noted that extreme family
cohesion (either enmeshment or disengagement) were associated with guilt-inducing
communication as well.
It is clear that communication within the families of survivors both about the
Holocaust and in general is a potential pathway for the transmission of trauma.
Dysfunctional methods of communication such as obsessive retelling of events,
complete silence, indirect, ambiguous communication as well as communication laced
with negative emotions that can be seen as guilt inducing have all been linked to
psychological problems in the children of survivors.
3.5. – Summary and Conclusions
The model presented at the end of Chapter Two has been expanded to
incorporate these family interaction variables. They have been bolded in the revised
version of the model presented overleaf in Figure 3.1. In contrast to the influential
psychological process variables, which are thought to influence the severity of
symptoms experienced, the family interaction/possible modes of transmission variables
© Janine Lurie-Beck 2007
60
are argued to mediate the relationships between ancestor and descendant variables. For
example, the relationship between Holocaust survivor scores on psychological impact
variables and child of survivor scores on psychological impact variables is hypothesised
to be mediated by the family interaction variables.
While the majority of literature on transmission of Holocaust trauma refers to
the transmission from survivors to their children, it is important to also remember
grandchildren of survivors. Numerous authors have argued it is important to include
assessments of grandchildren of survivors so that the parenting of the children of
survivors can be evaluated (Bar-On, 1995; Bar-On & Gilad, 1994; Mazor & Tal, 1996;
D. Weiss, 1988).
Indications of transmission from children of survivors to the
grandchildren have begun to appear in the literature (Bistritz, 1988; Brom et al., 2001;
Chaitin & Bar-On, 2002; Ryn, 1990) but research is still lacking (Chaitin, 2003).
3rd Generation
(Grand-children of
Survivors)
2nd Generation
(Children of Survivors)
1st Generation
(Survivors)
Psychological Impacts of
the Holocaust
Influential
Psychological
Processes
Depression
Anxiety
Paranoia
PTSD symptoms
Romantic Attachment
Dimensions
• Post-traumatic Growth
• World Assumptions
• Coping Strategies
•
•
•
•
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
• World Assumptions
• Coping Strategies
•
•
•
•
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
• World Assumptions
• Coping Strategies
•
•
•
•
•
Modes of Intergenerational
Transmission
of Trauma
• Parent-Child Attachment
• Family Cohesion
• Encouragement of
Independence
• General Family
Communication
• Communication about
Holocaust experiences
• Parent-Child Attachment
• Family Cohesion
• Encouragement of
Independence
• General Family
Communication
Figure 3.1. Addition of Trauma Transmission modes to the Preliminary Model of the Differential Impact of Holocaust
Trauma across Three Generations
© Janine Lurie-Beck 2007
61
Chapter Four – Demographic and Situational Differentials in the Impact of the
Holocaust on Survivors
I was 18 years old when everyone I loved was taken from me: my
home, my belief in people, my nationality, everything. I don’t believe
I ever recovered from these. I live under the shadow of the Holocaust.
Of course I have happy occasions, but even these times I felt sad
because my family was not with me.
A Holocaust Survivor
Chapter Two introduced the idea that there is evidence of differential post-Holocaust
adaptation relating to various demographic variables. Some of these have been directly
assessed by research and others have been left unexamined though the potential to
evaluate their impact has been present on many occasions.
The idea that there may be differences in adjustment among the Holocaust
survivor population (both in terms of their own psychological health and the way in
which they interact with their children) has been espoused by a number of researchers in
the field (Eitinger, 1969; Grubrich-Simitis, 1981; Halik et al., 1990; Kellerman, 1999).
Grubrich-Simitis (1981) notes that “there is no obligatory correlation between having
survived the concentration camps and the emergence of belated psychic aftereffects.”
While some survivors remain/ed deeply affected by their experiences many adjusted to
post-Holocaust life very well and lead productive lives (Halik et al., 1990).
Bistritz
(1988) suggests that a continuum of adjustment exists among survivors with varying
levels of adjustment possible although she doesn’t suggest what factors influence a
survivor’s position on this continuum.
McCann and Pearlman (1990) argue that the exploration of individual
differences in response to massive traumas, such as the Holocaust, should be an
important area of research within the trauma field. The suggestion that reasons for
differential adjustment should be explored within the Holocaust survivor population
was made as much as thirty years ago (for example Antonovsky et al., 1971) but largely
ignored. There has been an acknowledgement that the tendency of the majority of
research to treat Holocaust survivors as a homogenous group is limiting (L. Berger,
1988; Bistritz, 1988; Blank, 1996; Danieli, 1998; Kahana & Kahana, 2001; Marcus &
Rosenberg, 1988; McCann & Pearlman, 1990; Shmotkin & Lomranz, 1998). This
chapter seeks to delineate potential moderating demographic variables (and interactions
between them) and incorporate them into the proposed model of the differential impact
© Janine Lurie-Beck 2007
62
and transmission of Holocaust trauma which has been gradually built up over previous
chapters.
The specific Holocaust survivor demographic and situational variables to be
considered in this chapter are age during the Holocaust, time lapse since the Holocaust,
gender, country of origin, cultural differences, reason for persecution, nature of
Holocaust experiences, loss of family members, post-war settlement location and postwar resettlement issues. While these variables have been applied to the specific case of
the Nazi Holocaust, the majority of them appear in some form in either or both of the
two trauma models that have been used as a reference point for the current thesis:
namely Green et al.’s (1985) Working Model for the Processing of a Catastrophic Event
and Wilson’s (1989) Person-Environment Approach to Traumatic Stress Reactions.
Both theories refer to demographic characteristics, specific elements of the trauma itself
as well as aspects of the “recovery environment” as being influential in the severity of
traumatic reaction experienced by a victim/survivor.
4.1. – Age during the Holocaust
That there is, or at least might be, a difference in adjustment level among survivors of
differing ages was first raised very early on. For example, Niremberski (1946) noted
that children’s psychological well-being differed from that of adults in displaced
persons camps. Since then numerous researchers have argued for the influence of a
survivor’s age during the Holocaust on their subsequent adaptation (Brom et al., 2002;
Chaitin, 2003; Dasberg, 2001; Kahana & Kahana, 2001; Kellerman, 1999, 2001a,
2001b; Kuperstein, 1981; Matussek, 1975; Ornstein, 1981; Reijzer, 1995; Ryn, 1990;
Shanan, 1989; Steinberg, 1989; Suleiman, 2002; Tuteur, 1966).
Theorists have differed in their opinions as to the most vulnerable age group.
Most argue that a negative relationship between age and psychological symptomatology
exists in the Holocaust survivor population (Auerhahn & Laub, 1987; Baron et al.,
1993; Bower, 1994; Brom et al., 2002; Budick, 1985; M. Cohen, Brom, & Dasberg,
2001; Dasberg, 1987; Kahana & Kahana, 2001; Kellerman, 2001a; Mazor, Gampel,
Enright, & Orenstein, 1990; J.J. Sigal, 1998), while others argue for a positive
relationship (Kestenberg, 1990, 1993). Still a third group argues that adolescents were
the worst off with those younger and older in better psychological shape (Bower, 1994;
Budick, 1985; Felsen, 1998; Kuperstein, 1981; Marcus & Rosenberg, 1988; Suleiman,
2002). The arguments presented by these three camps will be addressed in turn.
Krell’s (1985) comments précis the main argument for why the youngest among
survivors would be most severely scarred by their experiences. These young survivors
© Janine Lurie-Beck 2007
63
were “too young to have partaken of a foundation for life, too traumatised to experience
a childhood and too preoccupied with survival to reflect on its impact.”
Firstly, in terms of foundations, Rustin (1988) suggests that the degree to which
Holocaust survivors were able to develop an empathic relationship with their parents
before being separated from them had implications for their own child-rearing
capabilities. Specifically, if the survivor was lacking a model of empathic parenting
from their parents they would be unable to respond appropriately to their own children
as they went through the “normative crises of childhood and adolescence”. The key
point here is that the younger the survivor was during the Holocaust, the less of a
chance they had to develop the empathic relationship with their parents. Therefore
younger survivors are envisaged to have had less empathy and greater difficulties with
their children at these crises points. However, some young survivors may have been
provided with a model of empathy from surrogate parental figures during the Holocaust.
Therefore this rule may not universally apply (Rustin, 1988).
Memories of pre-war family life were of great comfort to many while enduring
the traumas of the Holocaust (Rustin, 1988). According to Valent (1998), children from
as young as four years of age are able to maintain an image of their parents, fleshed out
by memories of their last moments together which they can remember forever. These
images are required for the development of a secure self (Kestenberg, 1985) or as
Moskowitz (1983) suggests are the nucleus of identity. Survivors old enough to clearly
remember this period and who had developed their own identity to some degree are
described as feeling more centred and self-reliant than younger survivors (Rustin,
1988).
In a qualitative assessment of a group of well-adjusted Holocaust survivors, Lee
(1988) had the following to say on the positive influence of positive childhood
memories:
The capacity to survive was grounded at least in part on early
experiences of a positive nurturing environment. Apparently, having
received love, warmth and respect from their parents had facilitated
their enduring of a terrible ordeal and their going on to living
meaningful lives in spite of the severe trauma they had experienced.
The child brought up in a home with warmth and nurturance with
experiences of love and security will have a reservoir of strength upon
which he or she can draw in times of need (p.76).
That the Holocaust afforded very little opportunity for normal childhood
activities is an obvious point. As Dasberg (2001) puts it “children were exploited and
deprived of normal patterns of schooling, games, friendships with other children, and
© Janine Lurie-Beck 2007
64
hope for the future.” More specifically the situation their parents found themselves in
precluded them from establishing an adequately safe and trusting environment that
children rely on them to provide (Dasberg, 2001; Macksoud et al., 1993; Rustin, 1988).
This is quite apart from the physical requirements of adequate shelter, nutrition and
hygiene which were also largely lacking (Dasberg, 2001). Macksoud et al. (1993)
contend that the fear/anxiety felt by children when their parents are unable to protect
them from harm can be very intense. The Nazis success in degrading their parents in
front of their children destroyed their image of protecting and loving caregivers
(Kestenberg & Kestenberg, 1990b)
Auerhahn and Laub (1987) argue that the nature of familial losses a younger
survivor experienced are more damaging in the long-term than the losses of older
survivors. Specifically they suggest that survivors who were children or adolescents
lost their family of origin which they argue has more serious implications for
attachments and parenting ability than adult survivors who lost their children. It is
assumed the underlying logic of this argument is that these younger survivors lost initial
and primary attachment figures. However, older survivors were likely to have lost their
parents and their children during the Holocaust which it could be argued would be
potentially more traumatising. People old enough to have grown up children were often
killed as they were considered too old to work and where therefore of no use (Chaitin &
Bar-On, 2002).
Having summarised arguments for a negative relationship between age and
symptomatology (i.e., that younger survivors are/were worse off than older survivors)
attention can now turn to points made for the opposing theory. The main stance among
theorists espousing this view is that youth brings with it the ability to recover quickly.
Kestenberg (1993) argued that once the child survivors went through the initial
physical recovery they were essentially better able to “bounce back” than adults. She
suggests that the knowledge that their whole life was ahead of them was a positive
influence on them. For adults, on the other hand, the treatment endured in the camps
was more psychologically damaging as it led to regression back to earlier stages of
development.
Adults regressed back to pregenital phases (within a Freudian
framework) because they were treated as if they were children.
According to
Kestenberg (1993) this additional regression factor meant that adults could not so
readily readjust to normal conditions.
Children may be viewed as more vulnerable but are also viewed as more
adaptable.
Kellerman (2001a) stated that “children are at the same time more
© Janine Lurie-Beck 2007
65
vulnerable and more malleable than adults.” In the immediate post war period
Niremberski (1946) remarked that children under eight years appeared to show no
symptoms of fear or anxiety and that while such symptoms were seen in the older
cohort of eight to sixteen year olds, their “readjustment was reasonably fast.” Matussek
(1975) noted that younger survivors were more socially integrated in their post-war
lives than older survivors.
Further support for the notion that younger survivors had better post-war
adjustment is found in a study by Eitinger (1973). A quick examination of the age
profile of the groups in this study reveals that the average age of survivors seemingly
well adjusted and in the workforce is noticeably younger than that of groups who had
come to the attention of health services. A later study conducted by Eaton, Sigal and
Weinfeld (1982) found the incidence levels of symptoms such as depression and anxiety
to increase with age in a group of male survivors.
The final viewpoint in relation to the impact of age on the impact of the
Holocaust is that adolescents represent the most vulnerable subgroup of survivors.
Budick (1985) explains that adolescents as a group could be considered more vulnerable
to Holocaust traumata because of the nature of the adolescent period. Even without the
backdrop of the Holocaust, adolescence is a period of emotional upheaval and
sensitivity which makes them more vulnerable to environmental changes than people
who have already successfully passed through this developmental stage (Budick, 1985).
Suleiman’s (2002) argument is couched in terms of cognitive development.
Specifically he separates child survivors into three basic groups based on their ability to
comprehend and deal with their situation. The first group aged up to 3 years he
considers too young to remember and unlikely to be affected to any large degree. The
second group, aged between approximately four and ten years are old enough to
remember but too young to understand. The third group, aged in the early adolescent
years of approximately 11 to 14 are considered the most vulnerable because they are old
enough to understand but too young to be responsible. By this age they have the
cognitive ability to understand the implications of their situation (Macksoud et al.,
1993). This age group were often forced to take on responsibilities and make choices at
a much younger age than they normally would have. In other words, they were forced
to take on adult roles and responsibilities prematurely (Macksoud et al., 1993;
Suleiman, 2002). The stress of this role for this age group is seen to have further
compounded their already traumatic situation (Suleiman, 2002).
© Janine Lurie-Beck 2007
66
It is worth noting that age had a significant influence on the likelihood of
survival in the first place. Many authors have noted that the young and old were often
sent to their deaths in the gas chambers immediately upon their arrival at concentration
camps (Chaitin & Bar-On, 2002; Nathan et al., 1964; Sternberg & Rosenbloom, 2000;
Sugar, 1999). Adolescents and younger adults were kept alive in order to perform
physical labour (Sugar, 1999).
Survival of young children did occur but was the
exception rather than the rule (Sternberg & Rosenbloom, 2000). Therefore the majority
of survivors would have been in their teens or twenties (Brody, 1999; Krell, 1997a;
Nathan et al., 1964; Ornstein, 1981)
Camp survivors are also mainly made up of those who came to the camps
towards the end of the war as very few survived for lengthy periods of time (i.e., years)
in camps. Young children had a better chance of surviving the war if they were in
hiding (Brody, 1999). This point therefore raises a potential confound between age and
the nature of Holocaust experiences. Support for this argument comes from a study
conducted with Bosnian refugees persecuted by Serbians in the 1990s. Weine et al.
(1995) found that, among a group of 20 Bosnian refugees, a positive relationship
between age and negative symptoms was no longer statistically significant when
traumatic exposure was partialled out. In other words, it was the amount of traumatic
exposure that varied with age and this was the variable impacting on number and
severity of symptoms.
In contrast to theoretical discourse concentrating on the relationship between age
and degree of psychological impact, there is also a sizeable amount of literature
referring to the relationship between age and the nature of psychological impact. For
example, Bensheim (1960 (in German), cited in Hafner, 1968) describes three very
different symptom patterns based on age during incarceration in concentration camps.
Survivors interned in camps between the ages of six and twelve were described as
suffering from “paroxymal affective reactions”, and displaying a fear and mistrust of the
environment. Survivors in their adolescent years (twelve to sixteen) also displayed fear,
however psychosomatic symptoms were dominant. Chronic depression appeared to be
the main symptom seen in survivors over 30 years of age (Bensheim (in German), 1960,
cited in Hafner, 1968).
Some symptoms are viewed as universally apparent irrespective of the age of the
survivor during the Holocaust while others display a clear relationship with age. For
example, consistent with Bensheim’s (1960 (in German), cited in Hafner, 1968)
© Janine Lurie-Beck 2007
67
idea that depression was a dominant symptom in survivors who were over 30, Hafner
(1968) notes that depression appeared to be more common among older survivors than
younger survivors, both from appraisal of his own sample and review of earlier
literature. In contrast, anxiety and paranoia levels seem to be more independent of age
(Hafner, 1968).
In their study of child survivors of the Holocaust, Keilson and
Sarphatie (1992) found that children aged up to 4 years during the Holocaust suffered
from what they termed “neurotic character development”, while those aged 11 to 14
when liberated commonly experienced anxiety and older teens more commonly
experienced depression.
Differences in relational factors are also postulated. Steinberg (1989) relates
differences in parenting experiences among survivors of varying ages. For example,
older survivors who lost their partner and perhaps children of their own are described as
becoming over-invested in their post war families while younger survivors who lost
their parents rather than children were more focussed on fears of abandonment
(Steinberg, 1989). Survivors who experienced the Holocaust in their adolescent years
were often seen to have difficulties relating to their children when they reached this
developmental phase (Kestenberg, 1985).
4.2. – Time Lapse since the Holocaust
Another issue of import is the impact of aging on survivor’s mental health. A number
of researchers have suggested that survivors’ symptomatology worsens with age (e.g.,
Dasberg, 1987; Joffe et al., 2003). In particular, some suggest that the process of life
review, that is dealt with as a person reaches their later years, is a negative process for
survivors as they have to review their Holocaust experiences as well as other life
experiences (Lomranz, 1995). The revision of these experiences brings the anguish of
the period and its losses back to the fore. The survivors are then confronted with the
resurfacing of issues related to incomplete mourning (Dasberg, 1987).
It has been suggested that there may be a curvilinear relationship between time
lapse since the Holocaust and intensity of symptoms. In other words, with time, the
symptoms gradually decrease in severity but then increase again when the life review
process occurs (Dasberg, 1987).
Joffe et al. (2003) state that, in their clinical
experience, symptoms suffered by Holocaust survivors appear to worsen with age.
Terno, Barak, Hadjez, Elizur and Szor (1998) note that exacerbation of symptoms such
as suicidal ideation and depression, appearance of paranoia and chronicity of
schizophrenia are often present in aging Holocaust survivors they have encountered.
Dasberg (1987) cites results of an unpublished survey at the Jerusalem Clinic for
© Janine Lurie-Beck 2007
68
Psychiatry of the Elderly in which 25% of new referrals between 1983 and 1986 were
Holocaust survivors. He notes that the majority of these new clients were experiencing
symptoms that appeared to be related to their Holocaust experiences (Dasberg, 1987).
Terno et al. (1998) also note a disproportionate number of Holocaust survivors among
their geriatric psychiatric patients than would be expected by their proportion in the
Israeli population (60% compared to 40%).
4.3. –Gender
Apart from age, gender is one of the most frequently mentioned demographic variables
in the Holocaust survivor research (Kahana & Kahana, 2001; Matussek, 1975; Tuteur,
1966) and the possible differential impact of gender was recognised very early
(Niremberski, 1946). While conflicting theories regarding the impact of gender of
adaptation to Holocaust trauma were found, on the whole it is thought that females are
affected more than males by their experiences.
However it is worth considering
arguments on both sides.
Baumel (1999) and Danieli (1982) have espoused two different reasons as to
why male survivors may appear to be worse off than females.
Baumel’s (1999)
argument relates to the way females developed bonds with fellow inmates during their
incarceration in camps while Danieli’s (1982) focus is on what could be described as the
“wounded pride of the patriarch”.
Danieli (1982) suggested that males see themselves as “protector and provider”.
She argues that the way survivors were treated in concentration camps led to feelings of
complete helplessness which seriously challenged this male self-image. Males not only
had to deal with the experiences themselves but with the knowledge that they were also
unable to “protect and provide” for their family. In a similar vein a number of other
authors have suggested that children of survivor fathers might be less well adjusted than
those with survivor mothers because knowing that one’s father was a Holocaust victim
might destroy the individual’s image of the powerful father figure. It has been argued
that the image of the demeaned mother is less threatening than the image of the
victimised, defenceless father (Baron et al., 1993; Kestenberg, 1980). However, it was
countered by Pines (1986) that children of survivor mothers should be more affected
because if the mother has been traumatised she is unable to provide a secure foundation
for her children right from infancy.
An alternative view which also espouses greater male vulnerability relates to the
relative safety of both genders among those in hiding. Several researchers have noted
that males/boys in hiding during the war were more at risk than females/girls because of
© Janine Lurie-Beck 2007
69
the Jewish custom of circumcision (Rosenbloom, 1988; Sternberg & Rosenbloom,
2000; Valent, 1998). In Europe, only Jewish males were circumcised (Rosenbloom,
1988; Sternberg & Rosenbloom, 2000). Thus the Jewishness of a captured male could
be established easily by the Nazis. Because of this, people were less willing to help a
male hide than a female (Reijzer, 1995; J.J. Sigal, 1998). Females had more chance of
convincing the Nazis that they were not Jewish than males and therefore of escaping
capture. Therefore among survivors who spent time in hiding, it could be argued that
the males were in a more anxiety provoking situation than the females. However, there
is also the point that males were more likely to be considered fit and strong enough for
slave labour than females which led to the survival of more males than females from the
camps (J.J. Sigal, 1998).
Baumel’s (1999) view point is based largely on the idea that women
incarcerated in camps actively sought to bond with their fellow camp inmates to form
substitute families and support networks to a greater extent than men. Baumel (1999)
contends that constructing these networks came easier for the women. Within these
relationships the women found comfort and nurturing that theoretically helped them to
deal with their surroundings (Baumel, 1999).
Ainsworth (1991) explains that this
tendency may have a genetic/evolutionary basis.
As the “physically weaker sex”,
females required support and protection from others more than the physically strong and
independent males.
Gilligan (1984) suggested that women had more of a need to establish these
relationships. Gilligan (1984) argued that for men mature relationships reinforce the
idea of separation and independence while women continue to value attachment to
others. This is seen as another reason why women forged stronger bonds more rapidly
in the camp environment. This notion is further reinforced by Matussek’s (1975)
findings that in the immediate post-war period female survivors were much more likely
to marry in haste than male survivors. The male survivors were more likely to try to reestablish themselves via work while females sought to do so by establishing a family.
However, this theorised need for attachment among women could also lead to more
negative psychological impacts when separations from families or surrogate families
occurred.
However, as stated at the start of this section, the predominant view is that
female survivors suffer more from post-Holocaust effects than males. Dasberg (1987),
Kahana and Kahana (2001) and Matussek (1975) have all argued that females are more
deeply affected.
© Janine Lurie-Beck 2007
70
A number of studies considering the incidence of mental health symptoms in
survivors found higher incidences of various symptoms and disorders among the
females in their samples. For example, Sigal and Weinfeld (1989), found that 65% of
the females in their sample (as opposed to 36% of the males) showed some evidence of
psychiatric impairment. Eaton, Sigal and Weinfeld (1982) found that 64% of females
suffered from four or more symptoms compared to 36% of males in their sample. In
their longitudinal health study, conducted in a neighbourhood of Jerusalem since 1969,
Levav and Abramson (1984) found that 73% of female camp survivors versus 57% of
male camp survivors suffered from some emotional distress.
It is also noteworthy that in the Levav and Abramson (1984) study, incidence of
emotional distress was higher among females in the control group. However, the
criteria for inclusion in this control group will have included many who suffered some
degree of Nazi persecution and who may have lost family members in the Holocaust.
Despite this control group contamination it begs the question of whether higher
incidence or greater severity of symptoms among females is unique to the survivor
population or is true of the general population as well. A study conducted by Yehuda,
Halligan and Bierer (2001) found a much higher incidence of anxiety among control
females than males in their community study adding further support to this notion. It
would appear that females are generally more likely to experience emotional distress
and that this is also reflected in the survivor population and is not unique to it (for
example Oltmanns & Emery, 1995).
4.4. – Country of Origin
The survivors’ country of origin has rarely been studied as a potential moderating factor
despite the fact that the nature and duration of a survivor’s experiences could vary
greatly depending on this (Brom et al., 2002; Kellerman, 2001b; Suleiman, 2002).
Persecution began as far back as 1933 for German nationals (Edelheit & Edelheit,
1994), with Austria and Czechoslovakia following just before the war started and other
countries added to the list as the German army advanced.
For this reason some survivors endured over 10 years of gradual worsening of
conditions, while others endured a much more rapid decline in conditions which lasted
for around 3 years (Shanan, 1989). Bower (1994) notes that Polish subjects in his study
were persecuted for an average of five and a half years compared to only one or two
years for Czechoslovakian and Hungarian subjects.
Eitinger (1973) notes similar
differences between his group of Norwegian survivors and his other group originating
from Eastern European countries such as Poland, Hungary and Czechoslovakia.
© Janine Lurie-Beck 2007
71
Moreover, some countries were viewed more negatively by the Nazis than others and
therefore their citizens were subject to different levels of respect. For example, the
Germans had a much more negative view of the Poles than of the French or the Dutch.
Eitinger (1981) notes that although Denmark was invaded in 1940, the Danish
government largely maintained control and so persecution of Jews only began in 1943
when the Germans took over political as well as military power. Countries occupied by
German allies rather than by Germany itself also enjoyed some initial respite. For
example, much of Greece was occupied by the Italians and the Bulgarians and the
Jewish communities there were largely safe until mid 1943 (Sugar, 1999).
Factor (1995) provides a detailed definition of a Holocaust survivor which
includes the dating of when persecution of Jews occurred in a number of countries. He
dates persecution as beginning when each country was occupied by the Nazis or when a
complicit government began enacting anti-semitic laws and pogroms on their behalf.
Specifically the dates are: from 1933 for Germany, from March 1938 for Austria, from
October 1938 for Czechoslovakia, from September 1939 for Poland, from April/May
1940 for Denmark, Norway, Belgium, the Netherlands, Luxembourg and France, and
from April 1941 for Bulgaria, Rumania, Hungary, Yugoslavia, Greece and
countries/areas formally part of the USSR.
While it would be natural to assume that greater traumatisation/traumatic
reactions would be seen among survivors from countries with longer durations of
persecution, Davidson (1980b) would argue that this is not necessarily the case.
Davidson (1980b) believed that ghettoisation was a “strengthening process” which
made ghetto survivors better able to adjust to concentration camp life than survivors
who went straight to the camps (as often occurred in countries invaded in later stages of
the war). Davidson (1980b) cites the example of the Greek Jews who he argues faired
far worse in terms of mortality rates in the camp system. This argument is akin to the
scientific anecdote of throwing a frog into cold water and gradually heating the water as
opposed to throwing the frog into boiling water from the beginning – the process of
acclimatisation may well have gone a long way in determining survival and post-war
psychological adjustment.
A survivor’s country of origin also determined their post-war settlement location
to some extent. As was discussed in Chapter Two, some countries were easier and/or
more palatable or safe to return to than others. This point more specifically applies to
Jewish survivors.
© Janine Lurie-Beck 2007
72
4.5. – Cultural Differences
Quite apart from the differing historical timelines that go with separate countries, a
related but equally as important a variable is that of cultural differences (Brom et al.,
2002). There is evidence of ethnic/cultural differences in the variables of interest
outlined in the model being developed for the current thesis. Cultural differences exist
in family cohesion, separation-individuation and communication styles which may well
explain discrepancies between survivor and control groups (Aleksandrowicz, 1973;
Almagor & Leon, 1989; Felsen, 1998; Porter, 1981; Rosen & Weltman, 1996).
Matching survivor and control groups on country of origin controls this potential
confound, however there has been insufficient examination of cultural differences as a
factor which may influence responses and adaptation to Holocaust trauma.
A comparison of survivors from different countries or regions of Europe would
be advantageous in addressing this issue. While many studies have been found to cite
the countries of origin of their Holocaust survivor subjects, few conduct a comparison
of survivor sub-groups based on this variable. This represents a missed opportunity to
explore the impact of country of origin on survivor well-being that could easily have
been performed.
Jewish families, and particularly Eastern European Jewish families, are viewed
as generally more enmeshed than other groups, with parents, particularly mothers, seen
as over-protective (Aleksandrowicz, 1973; Felsen, 1998; Porter, 1981; Rosen &
Weltman, 1996).
Guilt-inducing communication is also a characteristic of the
stereotype of the Yiddishe Mammeh (Giordano & McGoldrick, 1996; Kellerman,
2001b).
Therefore, non-survivor Jews might be seen to experience similar family
environments to descriptions of survivor families (D. Weiss, 1988). However it has
been argued that the Holocaust may have served to intensify the already over-protective
and enmeshed family environment (Felsen, 1998; Kellerman, 2001b; Porter, 1981).
Kellerman (2001d) suggests the stereotypic statements of the Yiddishe Mammeh, for
example those concerning eating enough food, take on “more desperate and anxious
overtones.”
4.6. – Reason for Survivor’s Persecution
While most people are aware that Jews were a major target group of the Nazis, there
were others who suffered at the hands of the Germans. While alienation, via the
removal of civil rights and ghettoisation, was unique to the Jews and Gypsies, people
could be interned in concentration camps for a number of reasons (Favaro et al., 1999).
Camp populations were made up of targeted ethnic groups such as Jews and Gypsies as
© Janine Lurie-Beck 2007
73
well as political prisoners (mainly of left wing persuasion such as communists and
socialists), criminals, homosexuals and members of religious groups such as the
Jehovah’s Witnesses and those opposed to the Nazi regime on religious grounds
(Schmolling, 1984; J.J. Sigal & Weinfeld, 1989).
The experiences of inmates (and also therefore chances of survival) could differ
a great deal depending on their reason for persecution (Kren, 1989). Jews and Gypsies
received the worst treatment by far and were among the first killed or “exterminated” in
large numbers (Hodgkins & Douglass, 1984).
While all inmates suffered, those who
were considered Aryan, or captured members of the armed forces, for example, were
given larger portions of food and were not subject to discipline as harsh as that
experienced by Jews and Gypsies (Bluhm, 1948). In addition, as some were targeted as
individuals for individual acts, they were able to return to an intact family and home
upon liberation (Eitinger, 1969; Favaro et al., 1999). As a case in point among a group
of Italian political prisoners who participated in a study conducted by Favaro et al.
(1999) all had returned to their homes after liberation and only a very small number had
lost any family members.
Survivors targeted because of their race or religion often endured many years of
persecution before being sent to camps. By contrast, political and other survivors were
sent to a camp as a form of punishment for a “crime” or act against the Nazi regime
(Favaro et al., 1999). As well as experiencing differential treatment and post-war
outcomes, camp inmates imprisoned for acts against the Nazi regime or for “crimes”
were in a better position to rationalise their imprisonment than those targeted solely for
their ethnic group or race (Favaro et al., 1999).
The vast majority of research conducted with Holocaust survivors has
concentrated on the Jewish survivor subgroup (Favaro et al., 1999). While this group is
no doubt the largest segment of the survivor population, it is clear there are other groups
that merit research attention as well. The few research studies that have considered the
reason for persecution have found that it may well have a differential impact on postwar adjustment (Favaro et al., 1999; Matussek, 1975). For example, Matussek (1975)
found survivors persecuted because of their race (i.e., Jewish survivors) more frequently
exhibited symptoms such as depression and anxiety and were less trusting than
survivors identified as political prisoners. Favaro et al. (1999) concluded that political
prisoners had experienced far less traumatic effects of their experiences than those
incarcerated because of their race or ethnicity.
© Janine Lurie-Beck 2007
74
4.7. – Nature of Holocaust Experiences
Every survivor’s experience of the Holocaust was different. Holocaust survival does
not just encompass enduring internment in a concentration camp (Brom et al., 2002;
Friedman, 1948).
There were the preceding years of increasing persecution and
isolation from community and ghettoisation, which should not be ignored as
traumatising events/circumstances (Shanan, 1989). Beyond that, while many went to
camps, some managed to survive in hiding, which may have involved living in a
confined hiding space with or without the aid of others to provide food, or living in the
guise of an Aryan with false papers.
Others managed to escape into forests and
survived with partisan and resistance groups.
Many survivors experienced a
combination of these. That there may have been a differential impact based on the
nature of the Holocaust traumas experienced has been acknowledged by many in theory
but assessed by relatively few empirically.
The general consensus has been that camp survivors are the most detrimentally
affected of all survivors (Baranowsky et al., 1998; Eaton et al., 1982; Favaro et al.,
1999; Friedman, 1948; Jucovy, 1989; Kahana & Kahana, 2001; Nathan et al., 1964).
Diagnoses of post-traumatic stress disorder are noted as higher among this subgroup of
survivors (Baranowsky et al., 1998; Favaro et al., 1999; Kahana & Kahana, 2001).
Further support for this notion is Finkelstein and Levy’s (2006) finding of a negative
relationship between amount of camp experience and comfort with disclosure, given
that comfort with disclosure is generally negatively related to symptomatology.
The duration of camp internment has also been cited as a potential influencing
factor on post-war adjustment (Antonovsky et al., 1971; Kahana & Kahana, 2001; Last,
1989). It is hypothesised that survivors who spent time with partisan and resistance
groups were aided psychologically by the knowledge that they were actively trying to
undermine the regime trying to persecute them (Favaro et al., 1999; Jucovy, 1992;
Matussek, 1975; Okner & Flaherty, 1988; Porter, 1981; Steinberg, 1989)
There are also a number of authors who contend that survivors who spent time
in hiding are, if nothing else, at least more anxious than survivors with other Holocaust
experiences (Rosenbloom, 1988; Yehuda, Schmeidler, Siever, Binder-Brynes, & Elkin,
1997). The constant threat of discovery and the necessary high degree of vigilance for
self-preservation were certainly highly anxiety provoking circumstances.
Sleep
deprivation was often also necessary, feeding into a circular anxiety provoking situation
(Maller, 1964). For those hiding in solitary confinement, isolation was another factor
(Baron et al., 1993). In one study, a group of survivors who had spent the war in hiding
© Janine Lurie-Beck 2007
75
had a much higher incidence of depression and anxiety than other groups of survivors,
although small samples sizes mean generalisation is not yet possible (Eaton et al.,
1982).
Amir and Lev-Wiesel (2001) raised the related issue of the impact of forgotten
identity. Many child survivors were placed in hiding by parents or loved ones at such
an early age that they cannot remember their original family and identity. Most were
given new identities (especially Christian identities for Jewish children). Amir and LevWiesel (2001) found that survivors who have not been able to uncover their original
identity are plagued by symptoms of greater severity and/or frequency than those who
can remember their identity.
Despite all the evidence to the contrary, some are still of the opinion that the
nature of Holocaust experiences has no bearing on survivors’ post-war psychological
well-being (Dasberg, 1987; Terno et al., 1998). These researchers tend to lean on the
argument that when the traumatic experiences are so extreme and extensive there is
little value in trying to sub-categorise the forms of trauma endured (Grubrich-Simitis,
1981; Kahana et al., 1989; Kahana & Kahana, 2001).
The assessment of the impact of the nature of Holocaust experiences has been
conducted in one of two ways. One way has been to group survivors according to their
experiences and check for differences between the resultant groups.
Traditional
groupings include camp survivors, those who spent at least some of the war in hiding
and members of resistance/partisan groups (Felsen, 1998). The second way has been to
assign a severity of trauma rating to each survivor research participant based on their
experiences. The derivation of such scores requires the researcher to make largely
subjective judgements as to which experiences can be considered more traumatic than
others. Felsen (1998) notes that, to date, there has been no agreement in relation to the
definitive quantification of the relative traumatisation that occurred as a result of these
differing experiences.
A number of examples were found of studies where attempts to construct a
hierarchy of traumatic experiences were made. In general, camp incarceration is viewed
as the most traumatic of experiences. For the most part, researchers cite no justification
for their reasoning.
Last and Klein (1984) gave the following severity ratings in relation to
Holocaust trauma: a weight of 1 was given to those living under Nazi occupation in
Europe, a weight of 2 to those confined in a ghetto, a weight of 3 assigned to being a
camp inmate and a weight of 4 denoted experiencing both ghetto and camp life. Bower
© Janine Lurie-Beck 2007
76
(1994) also used a rank ordering of severity although this was said to be based on
survivors’ views of the severity of the experiences.
They ranked survival of an
extermination or concentration camp as the most traumatic, followed by ghetto living
and finally by what he termed “illegal living”.
Lev-Wiesel and Amir (2000) recognised that research has tended to
automatically assume camp experience to be the worst of the possible predicaments that
a survivor may have found themselves in. In fact they note that a large amount of
research has followed an unwritten rule that only those who spent some time in a camp
should be referred to as survivors. Robinson et al. (1990) also espouse the view that
greater psychological impact was felt by those who survived a camp as opposed to those
who were in hiding or had joined the partisans. Further illustrating this point, Carmil
and Carel (1986) suggest that emotional distress within their sample of Holocaust
survivors would have been greater had they limited their sample to camp survivors to
the exclusion of those who endured other experiences such as hiding. This statement
implies that they assumed the camp survivors would have higher symptom levels than
survivors with other non-camp experiences.
The predominant explanation as to why survivors who were interned in camps
may not have fared as well as others relates to how active a role the survivor played in
their situation. For example, Lev-Wiesel and Amir (2000) suggest that survivors who
spent the war in hiding or as partisans felt more in control of their lives because they
were doing something active toward the goal of self-preservation. They argue that
doing something active to counter the continuous life threats experienced enabled them
to somewhat preserve their “self-confidence, self-control and self-appreciation” (Felsen,
1998). The nature of the survivor’s role in their traumatic experience (active versus
passive) is seen as one of the key determinants of post-traumatic recovery (B. L. Green,
1993) and is noted in both Green et al’s (1985) Working Model for the Processing of a
Catastrophic Event and Wilson’s (1989) Person-Environment Approach to Traumatic
Stress Reactions.
This active-passive dichotomy can be operationalised in one way within the
survivor population in terms of resistance fighters versus camp inmates. Resistance
fighters are the ultimate example of people who played an active role in their war-time
experiences while camp inmates are often described as having passively conformed
with the processes laid out which led them eventually to the camps. This active-passive
dichotomy may well be acted out within the parenting styles of these survivors. Danieli
(1982) suggested that camp survivors taught their children to mistrust others and
© Janine Lurie-Beck 2007
77
discouraged their attempts to establish autonomy while ex-resistance fighters
encouraged their children to “take charge” with their children struggling to live up to
their “parent-hero image.” In a separate study, Bauman (2003) found that children of
camp survivors perceived their parents (particularly their mother) as more mistrustful
than children of survivors with other non-camp Holocaust experiences.
4.8. – Loss/Survival of Family Members during the Holocaust
They told her to go to the left and us to go to the right. And when
mama started crying they said oh it is only for the night we segregated
the young from the old. And they said you are going to have a shower
and be put into the barracks, you will see your children in the morning
and in the morning there was the smoke going in the crematorium that
was the morning. I don’t want to cry again…
When I come home and nobody was there and nothing was there and
all I wanted was a photo, just one photo of my parents and my sister
and I couldn’t get it anywhere, which to this day hurts me very much.
Jozefa Lurie, a Polish Jewish camp survivor who was the sole
surviving member of her family
The extent of familial losses experienced by a Holocaust survivor has been
hypothesised as one of, if not the most important predictor of post-war well-being of
survivors and their children (Axelrod et al., 1980; L. Berger, 1988; Chodoff, 1963;
Fogelman & Savran, 1979; Hafner, 1968; Kuperstein, 1981; Matussek, 1975). Degree
of bereavement is also listed in both Green et al’s (1985) Working Model for the
Processing of a Catastrophic Event and Wilson’s (1989) Person-Environment Approach
to Traumatic Stress Reactions as an important factor in determining the impact of a
traumatic event. Axelrod et al. (1980) suggest a negative relationship between the
number of surviving relatives of a survivor and the severity of symptoms experienced
by their children. The loss of family members in such horrific and violent circumstances
is something that often lead to “survivor guilt” and great difficulties with mourning
(Hafner, 1968) and has been said to be at the centre of the survivors’ trauma (Shanan,
1989; Valent, 1995). The death of family members was noted as causing the most
severe pain among a group of 36 survivors of the Armenian Genocide (Kalayjian,
Shahinian, Gergerian, & Saraydarian, 1996). To lose children via a violent death during
the Holocaust is seen as the most traumatising event that could have occurred to a
survivor. The death of a child before its parents is unnatural and difficult to come to
terms with, the death of a child in violent circumstances almost insurmountable (Felsen,
1998; Grubrich-Simitis, 1981; Shanan, 1989).
Some perspective on the impact of this issue can be gained by a quick survey of
some research studies in which the number of relatives lost during the Holocaust is
© Janine Lurie-Beck 2007
78
noted. Among a sample of 157 concentration camp survivors Matussek (1975) reports
that close to 80% had lost at least one family member and that 70% had lost practically
all their family members. In Nathan et al’s (1964) group of 157 camp survivors, only
13% could say they had not lost any family member during the Holocaust. In Hafner’s
(1968) sample of 86 survivors, 72% lost both their parents, 41% lost all their siblings
and of the those who had been married 48% lost their spouse and among the parents
57% had lost their children. Bistritz (1988) found that 72% of 32 camp survivors and
65% of 26 survivors who had lived underground had lost all their family members.
As Kestenberg and Kestenberg (1990b) so evocatively put it: “By separating
parents from children, husbands from wives, and siblings from each other, the Nazis
systematically re-created conditions comparable to those of an infant abandoned by his
mother.” Green (1993) lists the nature and number of familial losses as a key aspect of
trauma experiences in determining how well a survivor recovers from their experience.
More specifically if a person endured their traumatic experiences with another family
member such as a parent or sibling it is predicted that they will have more positive
psychological outcomes (Dasberg, 1987; Kestenberg, 1990).
For example Kinzie,
Sack, Angell, Manson and Rath (1986) found that child survivors of the Cambodian
genocide were less likely to evidence post-traumatic stress disorder symptoms when
they had been with a family member during their traumatisation.
Wilson (1989)
included both level of bereavement and whether the survivor/victim endured the trauma
alone or with others as variables of import in his Person-Environment Approach to
Traumatic Stress Reactions.
Anna Freud and Burlingham (1943) found children (exposed to trauma in World
War II) in the care of their own mothers or a familiar mother substitute were not
psychologically devastated by wartime experiences, principally because parents could
maintain day-to-day care routines and project high morale. Prot (2000) adds to this
argument by stating that survivors often told her that they felt much safer when their
mother was with them. Survivors who had family members with them did not have to
contend with loneliness on top of everything else (Prot, 2000).
Finally, referring
specifically to children, Macksoud et al. (1993) cite evidence that suggests that children
fare better if they remain with their families, even if they will endure or witness more
traumatic events. However, children who had lost parents and families were also likely
to be taken under the wing of other adults (Bar-On et al., 1998).
Therefore the
“buffering” provided by parents was also often provided by substitute caregivers. The
younger the child the more likely this was to occur (Bar-On et al., 1998).
© Janine Lurie-Beck 2007
79
The impact of these losses was lessened to some extent among those who were
able to establish close bonds with fellow survivors and establish surrogate families.
That such ties would be sought is no surprise. As Ainsworth (1991) notes: “That
affectional bonds are especially likely to be formed under conditions of danger is only
to be expected since the attachment and caregiving systems are concerned with needing
and providing protection.”
Such support has often been cited by survivors as an invaluable aid to their
coping during the Holocaust (Kahana & Kahana, 2001; Rustin, 1988; Schmolling, 1984;
Shanan, 1989). Among a group of 219 former camp inmates, 20% cited camaraderie
with fellow inmates as contributing to their survival (Matussek, 1975). Another study
(n = 143) found 63% attributed survival to social support within the camp and 39% to
help from their family (Suedfeld, 2003).
Involvement in this type of surrogate family provided a degree of normalcy to
the survivors’ day to day lives. These groups provided survivors with a forum in which
it was safe to express their feelings, let out frustrations and find support and
encouragement (Ornstein, 1981). Such opportunities were not available to the isolated
survivor. A degree of familiarity with home was an additional feature that made such
groups such a therapeutic force.
Having people from the same home town or
neighbourhood around enabled the maintenance of links to pre-war life (Ornstein,
1981).
However, Terno et al. (1998) provide a hypothesis for why a sole survivor may
have had far more difficulties adjusting to post-Holocaust life than those for whom
some family members survived. They argue that the relationship between Holocaust
experiences and psychopathology might not be a direct one but an indirect one. They
suggest that it may not be the Holocaust experiences per se that lead to
psychopathological symptoms but the lack of support available to sole survivors in the
post-war period that did the most damage. In other words, when their symptoms
emerged after the Holocaust they were limited in the support they had to help deal with
them (Terno et al., 1998).
The loss of family members during the Holocaust is a trauma that is keenly felt
for the rest of the survivors’ life as well as that of their children and possibly even
grandchildren. This is because the death of so many family members means the loss of
ones’ extended family, the loss of the survivors’ childrens’ ability to relate and be with
their grandparents and possibly aunts and uncles and cousins and learn valuable
information about their heritage and ancestry (Chaitin, 2002; Goldwasser, 1986;
© Janine Lurie-Beck 2007
80
Grubrich-Simitis, 1981; Russell, 1982). “It wiped out the possibility of a full life cycle
of multigenerational families where daughters learn how to be mothers from their own
mothers, where grandchildren hear stories and get a sense of continuity from their
grandparents (Bar-On, 1995).” With no extended family, the survivors and their
children had to rely on each other which, while often resulting in a close-knit family
group, often led to extreme family enmeshment (Fogelman, 1998). It means that the
survivors often missed out on having a model of growing old before they grew old
themselves since they were robbed of the opportunity to see their parent’s age
Of course I have the picture in my mind but I would love to at least
have pictures to show my children that they did have grandparents and
an aunt but they had nobody and I had nobody. Do you realise that I
didn’t even have any body to write a letter and send a photo to say I
have nice children and they are nice looking and they are nice for
someone to be happy that I have children. There was no-one in the
world. That is why my husband meant so much to me because he was
all I had. He was everything to me, he was the mother and the father
and lover and the father of my children and my husband and the friend
and the protector – everything.
Jozefa Lurie, a sole survivor of her family
4.9. – Post-war Settlement Location
Holocaust survivors are scattered all over the world. While some survivors remained in
Europe, most survivors emigrated from Europe to Israel or to far away
countries/continents such as the United States, Canada, Australia, South Africa and
various countries in South America (Bistritz, 1988; Danieli, 1988). Obviously these
locations differ in many ways and some have argued that differences in adjustment
levels may relate to the choice of post-war settlement location (L. Berger, 1988;
Chodoff, 1963; Danieli, 1988; Felsen, 1998; van Ijzendoorn et al., 2003). Porter (1981)
suggested that the motivation to emigrate or remain in one’s own country was probably
related to the severity of initial symptomatology experienced by the survivor.
The locations chosen by survivors for their post-war settlement can be broken up
into three main categories. These categories are those who chose to remain in Europe,
those who chose to immigrate to Israel and those who chose to immigrate to other
continents such as North America or Australia. For those who chose to emigrate,
immigration experiences and difficulties assimilating into new societies and cultures
(often having to learn a new language) were interwoven with their long term recovery
process (de Silva, 1999; Solkoff, 1981; van Ijzendoorn et al., 2003).
It is also important to consider which survivors were more likely to return to
their own country or remain in Europe or move to Israel or to a completely different
© Janine Lurie-Beck 2007
81
continent. This is related to a number of factors including whether they had anyone to
return to and also the reception they received from their communities.
Political
prisoners were more likely to have families to return to than Jewish prisoners for
example.
A survivor’s choice of post-war settlement location had the ability to greatly
affect their recovery and psychological well being after the Holocaust.
Markedly
different experiences and receptions awaited survivors depending on where they chose
to settle after the war. Due to the differing nature of these post-war environments it is
not unreasonable to hypothesise that recovery processes may have differed by location.
Green (1993) argues that the way a survivor of trauma is treated by society after the fact
is of vital importance to their long term prognosis. The possible importance of this
issue in relation to Holocaust survivors has been recognised by several researchers as an
area requiring further assessment (Baron et al., 1993; L. Berger, 1988; Chodoff, 1963;
Danieli, 1988; Felsen, 1998; Kestenberg, 1990; Kestenberg & Kestenberg, 1990a; J.J.
Sigal & Weinfeld, 1989; van Ijzendoorn et al., 2003). Sigal and Weinfeld (1989)
recommended 18 years ago that a meta-analysis be conducted to determine if
differences in adaptation occurred as a result of post-war settlement location. Baron,
Reznikoff and Glenwick (1993) suggested a comparison of survivors who settled in
America versus those who settled in Israel is warranted.
Any differences found between survivors living in Israel versus those in Europe
or those in America or Australia cannot be automatically linked to differences in these
societies and the receptions they provided to the newly arrived survivor refugees. It
needs to be acknowledged that different types of survivors with different types of
experiences chose different post-war settlement locations and this potential confound
cannot be ruled out. For example is a survivor in Australia more anxious than a
survivor in Israel because of how they were received when they first emigrated and how
they continue to be treated or is it a function of how anxious they were in the first place
about a re-occurrence of the Holocaust that they chose to settle in a country so very far
away? However, the pros and cons of each post-war settlement location are many and
varied. Each will be considered in turn in the following sub-sections.
4.9.1. – Europe.
Survivors who settled in Europe remained in the area/region in which they suffered their
trauma (Fogelman, 1988). These survivors either returned to their country of origin or
settled in another European country but remained within relatively close proximity to
the region of their suffering. It should be noted that most camp survivors felt they could
© Janine Lurie-Beck 2007
82
not return to their homeland (Kren, 1989). However, by returning to their native areas,
these survivors did not have to go through the acculturation process that those who
decided to emigrate were forced to (Jucovy, 1989).
This group of survivors often lived in the midst of their former persecutors or at
least in the midst of bystanders who did nothing to ease their plight during the war
(Fogelman, 1988; Kahana & Kahana, 2001). Survivors residing in countries such as
Poland and Hungary also had to deal with the emergence of communism (Kahana &
Kahana, 2001).
In Germany, Fogelman (1988) argues that the government’s decision to pay
compensation to survivors increased anti-semitic and anti-survivor feelings in a country
concerned about getting its economy into a healthy state in the post-war period.
Survivors in Germany were acutely aware of this societal sentiment and this led many
to keep their problems hidden (Fogelman, 1988)
The decision to remain in Europe was dictated to some extent by the survivors’
country of origin and their belief in their ability to stay on. Survivors from some
countries were more readily accepted than others with anti-semitic feelings rife in some,
especially Eastern, European countries. Many Jewish survivors returned home to find
non-Jews living in their homes and unwilling to relinquish their ill-gotten gains to their
rightful owners (Reijzer, 1995).
The thought that a return to their homeland was possible was a buoying
influence recognised by Niremberski (1946) in his visits to displaced persons camps.
He noted that psychological symptoms appeared to be diminishing among those who
felt they would be able to return home and that there would be something to return
home to. He notes that this was least likely to be the case for Poles and Jewish
survivors who, for the most part, had lost all possessions and all or most of their family
members (Niremberski, 1946).
Survivors from these regions who decided to return to their homelands
maintained a very insular life.
According to Orwid, Domagalska-Kurdziel, and
Pietruszewski (1995), who studied a group of survivors who returned to their Polish
homeland, most married out of the Jewish faith and maintained strict secrecy about their
Holocaust experiences. They note that these survivors also kept their Jewish origins a
secret. Similar descriptions of survivors in Poland were presented by Prot (2000).
Haesler (1994) suggests that for many survivors who remained in Europe,
particularly in Germany, this was more of a chance event than a deliberate decision.
Many survivors wanting to emigrate to other continents waited a very long time in order
© Janine Lurie-Beck 2007
83
to obtain the necessary paperwork and permissions to do so. In the meantime, these
survivors had to find a way to support themselves while they waited and in so doing
often set themselves up in such a favourable working situation that it seemed better to
remain than to leave. It was rare for a survivor to make the decision to settle in
Germany with all of the negative feelings that this would bring to the surface. For some
however, the knowledge that they were re-establishing a Jewish community in Germany
was a powerful motivator in their decision to stay permanently (Haesler, 1994).
4.9.2. – Continents other than Europe.
A second group consists of survivors who made a concerted effort to move far away
from Europe and chose to immigrate to completely different continents. The most
common studies of these survivors are conducted in America, however studies
conducted with survivors who moved to Canada and Australia can also be included in
this group.
Unfortunately many of these survivors who ventured far afield to re-establish
their lives were not always received with open arms. Many found that people did not
want to hear about their Holocaust experiences. As a result they kept to themselves and
mainly socialised with other survivors (Bistritz, 1988).
The reception awaiting
survivors once successful in their emigration efforts differed greatly. In many countries
survivors were shunned – people did not want to hear about their experiences. The
thought of what had happened to survivors was too much to bear, they did not want to
be reminded of it (Danieli, 1988). Alternatively, people did not ask survivors about
their experiences for fear of bringing up hurtful memories for them (Danieli, 1988).
In discussions about Jewish survivors there is a distinction made between
survivors who settled in Israel and those who settled elsewhere, or the Diaspora
(Kaslow, 1995).
Survivors in the Diaspora were much more likely to experience
alienation based on their survivor status and so often chose to remain silent so as not to
antagonise people and risk the success of their assimilation (Goldwasser, 1986; Kaslow,
1995). This became known as the “conspiracy of silence” (Danieli, 1988) and it further
compounded survivors sense of isolation and loneliness and their mistrust of society
(Danieli, 1988).
Fogelman (1998) argues that many Jewish survivors in America were rejected
by the American Jewish community who felt guilt over their perceived lack of, or
inadequate amount, of help they were able to provide to save Jews during the war.
Fogelman (1998) also notes that there was a notable degree of anti-semitism in America
in the immediate post-war period.
© Janine Lurie-Beck 2007
This kind of cultural milieu was certainly not
84
conducive to a smooth transition period for the survivors.
Many settled in close
proximity to other survivors and soon survivor enclaves were apparent in many cities
such as Pittsburgh and New York and Montreal in Canada (Fogelman, 1998). A similar
phenomenon was seen in Melbourne in Australia.
A key factor for survivors who moved to these English speaking countries was
the immigration experience and all that it encompassed. This included not only having
to get used to the new surroundings and very different cultural settings but also for
many learning to speak English (Bar-On, 1995; Jucovy, 1989). The move from Europe
to the Americas or Australia was a dramatic one and for many represented sudden and
permanent severing from their European homelands.
4.9.3. – Israel.
The third group consists of survivors, specifically Jewish survivors, who chose to
immigrate to Israel.
These survivors made a decision to move to a location
fundamentally associated with the religious/cultural group that made them targets
during the war. They therefore were fostering a very strong link to their Judaism. In
addition to this, these survivors have lived in an environment of continued warfare,
which continues to this day with hostilities between Israelis and Palestinians.
Immigration to Israel has been considered both a help and a hindrance to postwar adjustment among survivors. Arguments for the former position centre around the
benefits of feeling part of the emerging Israeli state. The later position draws on factors
such as continued warfare and the Sabra myth (van Ijzendoorn et al., 2003). Each of
these positions will be addressed in turn.
An important point to note is that Jewish survivors in Israel suddenly found
themselves to be part of the majority group rather than a marginalised minority group
(Hodgkins & Douglass, 1984; J.J. Sigal & Weinfeld, 1989; Z. Solomon, 1998). This is
something unique to Israel, the Jewish state. Jews in every other country in the world
are in the minority rather than majority (Kaslow, 1995). Jews in Israel could therefore
recognise that anti-semitism was less likely to take place given their majority status
(Hodgkins & Douglass, 1984) and easily fit in (Newman, 1979; van Ijzendoorn et al.,
2003). The sheer number of survivors who settled in Israel also meant that survivors
were surrounded by others who had gone through similar circumstances and with whom
they had some camaraderie (Goldwasser, 1986; Kaslow, 1995).
Survivors in Israel have the opportunity to take part in collective mourning
rituals such as Yom Hoshoah (a national day of mourning for those who died in the
Holocaust) (Aleksandrowicz, 1973; Bergmann & Jucovy, 1990; Goldwasser, 1986;
© Janine Lurie-Beck 2007
85
Newman, 1979; Steinberg, 1989).
Yom Hoshoah (Day of the Holocaust) was
introduced by David Ben Gurion, Israel’s first Prime Minister (Fogelman, 1988).
Bergmann and Jucovy (1990) suggest that such events provided survivors and their
families with a forum in which it was safe to work through negative emotions
associated with the Holocaust – something that might otherwise have remained
unvented. The public acknowledgement of their traumatic experiences provides the
survivors with additional validation and support. It also provides them with a further
opportunity to memorialise their friends and loved ones who died in the Holocaust and
to some degree may satisfy the need to witness that some survivors experience (Krell,
1993). Wilson (1989) specifically cites the availability of cultural rituals for recovery as
an important factor of the post-trauma milieu in his Person-Environment Approach to
Traumatic Stress Reactions.
It is precisely because of the increased level of public awareness and
commemoration of the Holocaust in Israel that children of survivors who settled in
Israel may be more educated about the Holocaust than children of survivors who settled
elsewhere. There is a strong emphasis in Israeli education about the Holocaust and the
role of survivors in the context of the establishment of the Jewish state. Israeli children
take part in school trips to Holocaust memorials and museums (with the impressive and
comprehensive Yad Vashem museum and archive based in Jerusalem).
Therefore
children of survivors in Israel are educated to a great extent about the Holocaust and are
therefore not solely dependent on their survivor parents for this knowledge (Okner &
Flaherty, 1988).
It is intuitive to suggest that survivors who settled in Israel may be worse off
than survivors who settled elsewhere in a more peaceful country (Antonovsky et al.,
1971; Dasberg, 1987). However, some have theorised that despite the background of
hostilities survivors who settled in Israel were provided with opportunities to mourn and
release anger not afforded to those in other countries (Antonovsky et al., 1971).
Researchers such as Bistritz (1988) and Hass (1996) suggest that the numerous
wars with surrounding Arab nations provided survivors with an outlet for pent up anger
via participation in the military with victories in these conflicts providing the boost that
can only be provided by being on the winning side (Hass, 1996; Porter, 1981). In
addition, the establishment of Israel as a Jewish state represented the failure of the Nazi
goal to wipe out the Jews completely (Newman, 1979; Z. Solomon, 1998). However,
Hass (1996) is quick to point out that the possible negative effect of continuing to live at
war must not be ignored.
© Janine Lurie-Beck 2007
86
Participating in the establishment and growth of the state of Israel (Hass, 1996;
Silverman, 1987) afforded survivors an opportunity to develop positive self-images as
pioneers building a homeland and refuge for Jews (Goldwasser, 1986). They were able
to identify their own rebirth and rehabilitation with the rebirth and growth of the state
(Z. Solomon, 1998). As Bistritz (1988) says “the transition from victim to contributor
was greatly enhanced in Israel.”
The Israeli ‘kibbutz’ communal style of living is also seen as critical to the
prevention of survivors’ alienation (Bistritz, 1988; Goldwasser, 1986). Nadler and BenShushan’s (1989) study findings tend to support this notion with survivors living in a
kibbutz found to be more emotionally stable, less tense, insecure and suspicious.
Survivors were more likely to receive much needed support and guidance with things
such as parenting in such a setting. However, Hass (1996) suggests that kibbutz living
may have actually been detrimental, arguing that it may be likened to ghettoisation.
There are numerous relatively convincing arguments that settlement in Israel
was more negative than positive for survivors. Holocaust survivors were sometimes
viewed as passively allowing themselves to be tortured and murdered by the Nazis
rather than attempting to fight (Bar-On, 1995; Bar-On et al., 1998; van Ijzendoorn et al.,
2003). This notion is inconsistent with the heroic Sabra ideal of strength and courage in
the face of adversity (Bar-On et al., 1998; Chaitin, 2002; Kellerman, 2006a). Posttraumatic weakness was also inconsistent with the Sabra and seen as detrimental to the
establishment of a strong state of Israel (Z. Solomon, 1998). That this ideal was quite
ingrained is supported by the publication of a pamphlet by the Israeli Ministry of
Education in the 1950s entitled “Like Sheep to Slaughter” which presented such
arguments in print (Bar-On, 1995). It is not surprising therefore that initially only those
who had actively fought in partisan groups or who had participated in ghetto uprisings
were seen worthy of commemoration.
The most obvious negative point is the amount of conflicts between Israel and
its Arab neighbours in the post-war period. First there was the War of Independence,
followed by the Sinai War, the Six Day War, the Yom Kippur War, war and terrorism
associated with conflict with Lebanon and numerous intafadas with the Palestinians
(Lomranz, 1995). While the political and military instability of Israel has been used as
a positive target for pent up anger over Holocaust experiences, it is doubtful whether
this argument can be stretched for the length of time that such circumstances have
existed in the Middle East.
© Janine Lurie-Beck 2007
87
Apart from the continued conflict in the area, there are other potential negative
impacts. The climate in Israel is quite hot at times (Matussek, 1975) and certainly much
less comfortable than a European climate. Also, as a newly established state, Israel was
poor in resources and economic stability (Matussek, 1975).
4.10. – Amount of Tme before Resettlement
It is important to reflect on the immediate post-war period during which time many
survivors spent extended periods of time in displaced persons camps awaiting a chance
to emigrate.
It has been suggested that this period further compounded their
traumatisation (Porter, 1981; Solkoff, 1981) as they “were made to feel over and over
again that they were unwanted – they were unwanted when they tried to return home
and unwanted when they tried to emigrate to a new country (Matussek, 1975).” Terry
(1984) added that despair stemmed from the disappointment with the way they were
received after their liberation. Jewish survivors found it very difficult to immigrate to
Israel prior to 1948 when the region was under British control. Countries all over the
world maintained tight restrictions on the number of refugees/displaced persons they
took in. Friedman (1948) noted that survivors who were able to gain relatively quick
access to these countries were able to re-establish a normal life quite quickly and
regained “their balance and health” whereas those who were left to languish in the
displaced persons camps suffered from increasing anxiety and aggression. Part of the
reason for this is that many displaced persons camps were quite regimented and were
experienced by many survivors as a return to the concentration camp lifestyle
(Friedman, 1948, 1949).
4.11. – Summary and Conclusions
This chapter has summarised the main arguments for the moderating impact of
demographic and situational variables on the impact of the Holocaust on its survivors.
While there is general agreement on the impact of some (for example gender), for the
most part there is disagreement as to the impact or direction of influence of most of the
variables discussed. Obviously, more clarity in the literature as to the impact of these
variables and their relative importance would be beneficial. This is one of the key
things the current thesis aims to address.
The preliminary Model of Differential Impact of Holocaust Trauma across
Three Generations is further augmented in Figure 4.1 to take into account the
demographic variables discussed in this chapter. The survivor demographic variables
added to the model have been presented in bolded typeface. They have been included
with reference both to the survivor generation as well as descendant generations. It is
© Janine Lurie-Beck 2007
88
felt that differences among survivors related to these demographic variables will
reverberate through the generations by virtue of differences in not only symptom levels
but parenting and familial patterns. Chapter Five goes on to discuss demographic
variables that are intrinsic to the descendant generations which may lead to differential
symptom
levels/trauma
© Janine Lurie-Beck 2007
transmission
among
survivor
descendants.
89
Psychological Impacts of the
Holocaust
Depression
Anxiety
Paranoia
PTSD symptoms
Romantic Attachment
Dimensions
• Post-traumatic Growth
3rd Generation
(Grand-children of Survivors)
2nd Generation
(Children of Survivors)
1st Generation
(Survivors)
•
•
•
•
•
•
•
•
•
•
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
•
•
•
•
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
Influential Psychological
Processes
Modes of Intergenerational Transmission
of Trauma
• World Assumptions
• Coping Strategies
• World Assumptions
• Coping Strategies
•
•
•
•
•
Parent-Child Attachment
Family Cohesion
Encouragement of Independence
General Family Communication
Communication about Holocaust experiences
• World Assumptions
• Coping Strategies
•
•
•
•
Parent-Child Attachment
Family Cohesion
Encouragement of Independence
General Family Communication
Demographic Moderators
Holocaust Survivor Generation
•
•
•
•
•
•
•
•
•
Age during the Holocaust
Time lapse since the Holocaust
Gender
Type/nature of Holocaust experiences
Reason for persecution
Loss of family
Country of origin
Post-war settlement location
Length of time before resettlement/time spent in displaced
persons camps
•
•
•
•
•
•
•
•
•
Age during the Holocaust
Time lapse since the Holocaust
Gender
Type/nature of Holocaust experiences
Reason for persecution
Loss of family
Country of origin
Post-war settlement location
Length of time before resettlement/time spent in displaced
persons camps
•
•
•
•
•
•
•
•
•
Age during the Holocaust
Time lapse since the Holocaust
Gender
Type/nature of Holocaust experiences
Reason for persecution
Loss of family
Country of origin
Post-war settlement location
Length of time before resettlement/time spent in displaced
persons camps
Figure 4.1. Addition of Holocaust survivor demographic variables to the Preliminary Model of the Differential Impact of Holocaust Trauma across Three Generations
© Janine Lurie-Beck 2007
90
Chapter Five – Demographic and Situational Differentials in the Impact of the Holocaust
on Descendants of Survivors
The current chapter summarises the demographic and situational variables that may have
influenced the nature of symptom presentation among descendants of survivors.
It should
be noted that all the demographic and situational variables that apply to survivors are also
argued to impact on subsequent generations. In other words, demographic subgroups of
survivors who are the most affected by their Holocaust experiences will most likely have
children who display the most or highest levels of symptoms also (Goldwasser, 1986;
Grubrich-Simitis, 1981; M. Weiss & Weiss, 2000; Yehuda, Schmeidler, Giller et al., 1998).
This is because the most traumatised survivors will have been the most ill-equipped for the
parenting role and the transmission processes discussed in Chapter Three will have been
highly present when raising their children. For the sake of brevity it was considered
unnecessary to revisit each survivor demographic variable discussed in Chapter Four with
regards to its continuing influence on the descendants of survivors in this chapter.
Therefore this chapter summarises the additional demographic variables that apply uniquely
to children and grandchildren of survivors with the understanding that survivor
demographics also play a role in determining the psychological health of their descendants.
5.1. – Children of Holocaust Survivor/s
Among children of survivors there are several demographic variables that are argued to
impact their psychological and inter-personal well-being in adulthood. These are: the delay
between the end of the war and their birth, whether they were born before or after their
survivor parents’ emigration from Europe (if this occurred), whether they have one or two
survivor parents, their gender and their birth order.
5.1.1. – Gender
The existence of potential gender differences within the children of survivor population has
been posited by a number of researchers. Male and female children are said to react
differentially to their survivor parents and their experiences (Solkoff, 1992b). Sons and
daughters of survivors may also interpret their survivor parents’ communication differently
(Felsen, 1998; Kellerman, 2001b; Rieck, n.d.-b).
In general, as with survivors, it is females that are said to evidence higher levels of
vicarious traumatisation or are considered more vulnerable to it occurring (Vogel, 1994).
Vogel (1994) argues that females are more vulnerable because the development of their
© Janine Lurie-Beck 2007
91
sense of self or identity is strongly related to their ability to “connect and feel alike” to
there care-givers (predominantly their mothers) while for males the focus is more on
creating differentiating, as opposed to osmotic, boundaries.
There is also discussion of gender differences in nature as well as severity of
symptoms or reactions to survivor parents’ trauma. For example, Felsen (1998) found that
sons of survivors react to stressful family environments with “introjective characteristics”
while daughters evidence more “anaclictic traits”.
Given that gender is such a readily assessable variable, it is striking to note that in
his review of the literature, Solkoff (1992a) found few examples of statistical analysis of
gender differences in children of survivor research. This is despite the fact that the majority
of studies were conducted with mixed gender samples and so the opportunity for such
analysis was clearly available.
5.1.2. – One versus Two Survivor Parents
It is commonly held that children with two survivor parents will have more psychological
problems than children with only one survivor parent (Grubrich-Simitis, 1981). However,
while the majority of conjecture has concurred with this assertion, there have been a small
number who have argued for the opposite, that having one survivor parent and one nonsurvivor parent is more detrimental to psychological health.
The arguments for greater negative outcomes for children whose parents are both
survivors are numerous. Davidson (1980a) suggests that a “dyadic victim unit” forms in
which each survivors’ fears and anxieties are reinforced by the similarly held fears and
anxieties of their survivor partner. Therefore, the whole may be more than the sum of the
parts. While having one survivor parent may relate to some dysfunctional or maladaptive
family interactional patterns (as compared to families with no survivor parents), it has been
argued that the non-survivor parent helps to mitigate these negative effects to a certain
degree (Gay et al., 1974; J.J. Sigal & Weinfeld, 1987).
Survivor “dyads” often formed very shortly after the end of the war, in many cases
whilst the survivors where still in displaced persons camps. Many felt that only a partner
who had also experienced the horrors of the Holocaust would be able to understand them
(Niederland, 1988).
In such cases, often their survivorship was the only thing these
survivor couples had in common, the realisation of which then led to unhappiness within
the marriage (S. Davidson, 1980a). However, just as a sense of loneliness and isolation had
often led to such hasty unions, the fear of this loneliness and isolation frequently led to an
© Janine Lurie-Beck 2007
92
unwillingness to separate (Bistritz, 1988; S. Davidson, 1980a).
Growing up in an
environment of an unhappy marriage is not overly conducive to the happiness of the
children of these marriages.
However, the argument for why having both survivor parents may actually be better
for children of survivors also uses the common background of survivorship as part of its
supporting evidence. Gay, Fuchs and Blittner (1974) contend that families created as the
result of a marriage between two survivors may be characterised by more harmony and
stability and less hostility because of an understanding of survivor partners’ backgrounds.
They argue on the basis of marital harmony literature that for couples made up of two
people from very different backgrounds (as would likely be the case in a survivor-nonsurvivor union), their differences serve to create a potentially ever-widening gap which
creates family unit weakness and discordance.
Given such an argument, children of
survivors would be more likely to live in an unhappy family environment if only one parent
was/is a survivor.
Van Ijzendoorn, Bakermans-Kranenburg and Sag-Schwartz (2003) found support
for the notion that having two survivor parents is worse than having just one in a metaanalysis of 32 studies conducted with studies assessing psychological well-being of
children of survivors. An effect size of 0.57 was obtained from the 13 studies in the two
survivor parent category compared to 0.09 from the 19 studies in the one survivor parent
category. However their assessment of this issue was not as valid as it could have been.
Studies were placed in the two survivor parent category if more than 75% of the sample had
two survivor parents with the remaining studies placed in the one survivor parent category.
This is by no means a pure comparison of the impact of having two versus one survivor
parent. A preferable method would have been to conduct a correlation between study effect
sizes and the percentage of the sample with two survivor parents. Merely dividing the
studies based on the 75% cut-off criteria meant that some studies in the two survivor parent
category would have included participants with only one survivor parent while large
proportions of study samples included in the one survivor parent category could have had
two survivor parents (indeed up to 74% of those samples). The more methodologically
sound approach of correlating study effect sizes with the percentage of the sample with two
survivor parents is used in the current thesis (see Chapter Nine, Section 9.2.1).
© Janine Lurie-Beck 2007
93
5.1.3. – Birth Order
First born and only children are viewed as being the most vulnerable of the children of
survivor group when considering birth order (S. Davidson, 1980a; Grubrich-Simitis, 1981;
Newman, 1979; Porter, 1981). It is thought that the only child with survivor parents is
particularly susceptible as they are the sole repository of the transmitted traumas of their
parents (Newman, 1979). Children in other birth-order positions are exposed to these to a
lesser extent because there are more children to mitigate the effect (Baron et al., 1993).
First born children are a physical representation of the recreation of families so
desired by Holocaust survivors (Porter, 1981). These children may be seen as resurrections
of lost loved ones, or so-called “memorial candles.” All the survivor’s hopes for the future
are concentrated on this child more so, it is argued, than subsequent children (Porter, 1981).
This leads to a large amount of pressure on this child to succeed compared to subsequent
children (Porter, 1981).
Other results/research on first-born children of survivors showed that these children
were more seriously affected by parents having been survivors than children in other birthorder positions because of the nature of the relationship between parents and first-born
children, namely, greater time spent alone with their parents, which would increase the
exposure to the survivor syndrome (Baron et al., 1993).
Arguments for the vulnerability of first born or early birth order children often
touch on a separate but related issue – that of the time lapse between the end of the war and
the birth of the children of survivors. As Grubrich-Simitis (1981) so aptly puts it: “This
enhanced vulnerability of firstborn children applies especially to those born very soon after
liberation, that is, where the birth of a new child was part of a manic attempt by the parents
to reconstitute the lost family and where this occurred before a certain level of psychic
reintegration – especially in the mother – could possibly have been reached (p. 431).” This
issue of the time lapse between the end of the war or survivor parent liberation and its
influence on the psychological health of children borne to survivors is discussed in the next
section.
5.1.4. – Length of Time between the End of the War and the Birth of Children
In Davidson’s (1980a) experiences with clients in Israel, an obvious relationship emerged
between the symptom severity of children of survivors and when they were born. He noted
that the majority of children of survivors, whom he saw in treatment, were born in the
immediate post-war period.
© Janine Lurie-Beck 2007
While children born after a longer delay were also
94
encountered, Davidson (1980a) describes their presentation as “less severe and of a more
subtle psychological nature.”
Russell (1982) concurred with Davidson’s (1980a)
experiences stating that these later born children evidenced “much less damaging aftereffects or were symptom-free.”
Oliner (1990) observed that a notable proportion of
children of survivors seeking help from mental health professionals were born in 1946. In
other words, directly after the war and often while the survivor parents were still in
displaced persons camps (Oliner, 1990). Many children were born in these camps, with
survivors conceiving as soon as their physical recovery allowed (Danieli, 1988). While the
general consensus has been that children born in the immediate post-war years are the most
vulnerable subgroup, there have been differing views. For example, Mazor and Tal (1996)
state that, in their experience, similar symptoms and interpersonal behaviour patterns were
seen both in children born soon after the war and those born much later.
In the immediate post-war period many survivors found themselves essentially
alone. Understandably, many survivors married and set up families in haste to ease the
loneliness they felt (Barocas & Barocas, 1980; Bistritz, 1988; S. Davidson, 1980a;
Grubrich-Simitis, 1981; Newman, 1979; Oliner, 1990; Porter, 1981; Shoshan, 1989).
Danieli (1988) has referred to such marriages as “marriages of despair” where the dominant
motivation for marriage was mutual loneliness and shared suffering (Freyberg, 1980).
Children were born soon after to further ease the loneliness and to replace lost family
members (Grubrich-Simitis, 1981): as so evocatively demonstrated in Figure 5.1.
Figure 5.1. Female Holocaust survivors with their babies born in a displaced persons camp. Florence, 1946.
Reprinted by permission from Elaine Zaks, whose mother is on the far right holding her brother as a baby.
© Janine Lurie-Beck 2007
95
The suggestion that improvement with delay is actually reflective of the degree to
which survivor parents had processed or resolved their mourning and integrated their
trauma has been made by many (S. Davidson, 1980a; Grubrich-Simitis, 1981; Oliner, 1990;
Russell, 1982). The psychodynamic viewpoint is that when insufficient time was allowed
for the survivor parents to resolve these issues, their working through process was played
out in their interactions and relationships with their children (S. Davidson, 1980a;
Grubrich-Simitis, 1981; Oliner, 1990). Bar-On et al. (1998) were quick to point out that it
is the lack of resolution, or mourning, that is the problematic issue not the loss or trauma
itself. In fact, if the parent has successfully resolved these issues prior to child-rearing,
often minimal effect has been noted.
Rubin (1983 (conference presentation), cited in
Hogman, 1985) stated that when mourning is resolved, memories of the lost loved one can
act as a source of warmth and happiness, whereas unresolved mourning brings negative
emotions with it. Further support for this notion can be found in the results of Ainsworth
and Eichberg’s (1991) study of the influence of maternal unresolved mourning on
children’s attachment with a non-Holocaust related sample. They compared children of
mothers who had lost a significant attachment figure and had resolved their mourning and
those who had not resolved their mourning. Ainsworth and Eichberg (1991) found that the
children with mothers who had resolved their mourning did not differ statistically
significantly in incidence of insecure attachment from a control group of children whose
mothers had not experienced the death of a loved one. However, the children of mothers
with unresolved mourning had a much higher incidence of insecure attachment types. This
finding serves to further highlight the message that it is not necessarily the experience of
the death of a loved one that is influential but the success with which this knowledge is
coped, or has been dealt with, that is key.
Unresolved grief and mourning have been commonly associated with the
experiences of Holocaust survivors. Many survivors lost not just one but many relatives in
horrific circumstances. Moreover, the circumstances in which they found themselves at the
time (for example in a concentration camp), precluded involvement in traditional aids and
rituals associated with mourning such as funerals and open expressions of grief (Barocas &
Barocas, 1980; B. B. Cohen, 1991; Freyberg, 1980). A number of authors have suggested
that the process of mourning was therefore suspended among survivors until after the war
ended. However, some survivors were not able to resolve their grief even in the post war
period. During the war, they were preoccupied with their mere survival, and then, in the
© Janine Lurie-Beck 2007
96
immediate post war years, they were preoccupied with re-establishing some semblance of a
normal life, finding employment and often learning new languages (Danieli, 1988;
Freyberg, 1980).
It is acknowledged that when children are born before the mourning process is
finished or resolved, it can have significant effects on the way the children are brought up
(S. Davidson, 1980a; Kuperstein, 1981). Indeed, the memories of the unmourned relatives
cast a shadow on the new family (S. Davidson, 1980a). In fact, some have argued that
rather than having a positive impact on their recovery, interactions with their children may
have further re-opened wounds in relation to past losses (Newman, 1979).
Ornstein (1981) validly queried if “to consider parenting as a potentially restorative
is risky, for at what price to the children does such recovery in the parents occur?” (p. 146).
Barocas and Barocas (1980) suggested that, while the majority of children of survivors born
in this early period were taken care of quite well in a physical sense, they can be described
as emotionally deprived. “There was considerable emotional deprivation due to the greatly
limited and restrictive affective resources, the impaired object relations and the griefstricken pre-occupation of the parents” (Barocas & Barocas, 1980, p. 8). Parents, preoccupied with their grief over lost relatives, were often emotionally unavailable to their
children (Barocas & Barocas, 1980) or emotionally absent (Barocas & Barocas, 1980;
Chayes, 1987; Kuperstein, 1981).
Children born to parents who had achieved insufficient resolution and integration of
their Holocaust experiences have been seen as more vulnerable to emotional problems
(Grubrich-Simitis, 1981; Oliner, 1990). According to Davidson (1980a), children born in
the immediate aftermath of the Holocaust became “intensely involved by their parents in
their early attempts at recovery.” (p. 20). Children born after longer delays did not get
caught up in their parents’ working through process, as their parents had “achieved a
greater degree of integration of the massive trauma” in the intervening years.
5.1.5. – Birth Before or After Parental Emigration
There is a second potential explanation for the seemingly increased levels of emotional
disturbance in children born soon after the cessation of hostilities. As Fogelman (1998)
pointed out, the experiences of these children were not only affected by their immediate
family, but also by the social climate in which they grew up.
The children born
immediately after the war were often born in displaced persons or transit camps and in
addition they may have experienced many moves between these camps. They may also
© Janine Lurie-Beck 2007
97
have accompanied their survivor parents on heart-breakingly fruitless searches for
surviving relatives (B. B. Cohen, 1991). Furthermore, their parents may well have still
been recovering physically from their ordeals and so may have been plagued by ill health
(Freyberg, 1980).
Beyond these experiences was the huge upheaval of migration to new countries. In
these new countries, their parents had to start again in establishing their lives, and their
adjustment was further complicated by the need to learn a new language and attempts at
assimilating into a culture/society different from what they had known (Fogelman, 1998;
Freyberg, 1980). The early experiences of children born in these circumstances compared
to those born after their parents had established themselves were clearly quite different
(Krell et al., 2004; Levav, Kohn, & Schwartz, 1998; Shoshan, 1989).
Levav, Kohn and
Schwartz (1998) suggested that given this disparity, children of survivors born in these two
different periods should be differentiated in research.
5.2. – Grandchildren of Holocaust Survivor/s
Just as the children of the most affected survivors are expected to be the most affected
themselves, so too are the grandchildren of those most affected. Again differences in
psychological well-being are expected based on gender and the number of Holocaust
survivor relatives.
As with the children of survivors (and for the same reasons noted above), gender
and birth order are hypothesised to be influential variables in the psychological well-being
of grandchildren of survivors. Research into grandchildren of survivors is a very recent
addition to the literature on the inter-generational impacts of the Holocaust. A handful of
studies have considered demographic variations in this population but more research is
needed into this issue.
5.3. – Summary and Conclusions
The model of differential impact of Holocaust trauma across three generations which has
been gradually forming over the previous three chapters is further amended here (see
Figure 5.2) to take into account the demographic variables considered in this chapter. As
can be seen child of survivor demographic variables have been listed against both the child
of survivor and grandchild of survivor generations.
As with the survivor generation
demographic variables, it is argued that these demographic variables potentially influence
not only the child of survivor generation themselves but also their children (the grandchild
© Janine Lurie-Beck 2007
98
of survivor generation). Once again the variables added to the model in this chapter have
been presented in bold typeface.
The addition of the descendant demographic variables completes the preliminary
version of the Model of the Differential Impact of Holocaust Trauma across Three
Generations that has been built based on a review of the theoretical and anecdotal
literature.
With the preliminary model built it is necessary to test it in some way. It was
decided that a systematic review of the existing empirical data in the literature regarding
Holocaust survivors and descendants would be cross-referenced with this model to
determine which if any areas of the model have already been adequately assessed by
previous research. In order to obtain a more objective synthesis of the existing data than
might be afforded by a traditional narrative review it was decided to use meta-analytic
techniques. This meta-analytic investigation is reported in Section B of this thesis. The
meta-analytic techniques used in this process are described in detail in Chapter Six,
including a more detailed justification for the use of meta-analysis. Chapters Seven to Nine
report the results of the meta-analyses conducted. The final chapter in Section B, Chapter
Ten, revisits the model in light of the evidence obtained from the meta-analyses.
© Janine Lurie-Beck 2007
99
Influential
Psychological
Processes
Depression
Anxiety
Paranoia
PTSD symptoms
Romantic
Attachment
Dimensions
• Post-traumatic
Growth
• World
Assumptions
• Coping Strategies
•
•
•
•
•
1st Generation
(Survivors)
2nd Generation
(Children of Survivors)
3rd Generation
(Grand-children of Survivors)
Psychological
Impacts of the
Holocaust
•
•
•
•
•
•
•
•
•
Depression
Anxiety
Paranoia
Romantic
Attachment
Dimensions
Depression
Anxiety
Paranoia
Romantic
Attachment
Dimensions
Modes of Intergenerational
Transmission
of Trauma
• World
Assumptions
• Coping Strategies
• Parent-Child Attachment
• Family Cohesion
• Encouragement of
Independence
• General Family Communication
• Communication about
Holocaust experiences
• World
Assumptions
• Coping Strategies
• Parent-Child Attachment
• Family Cohesion
• Encouragement of
Independence
• General Family
Communication
Demographic Moderators
Holocaust
Survivor Generation
Children of Survivor
Generation
•
•
•
•
•
•
•
•
•
Age during the Holocaust
Time lapse since the Holocaust
Gender
Type/nature of Holocaust experiences
Reason for persecution
Loss of family
Country of origin
Post-war settlement location
Length of time before resettlement/time spent in
displaced persons camps
•
•
•
•
•
•
•
•
•
Age during the Holocaust
Time lapse since the Holocaust
Gender
Type/nature of Holocaust experiences
Reason for persecution
Loss of family
Country of origin
Post-war settlement location
Length of time before resettlement/time spent in
displaced persons camps
• Number of survivor
parents
• Delay between the
end of the war and
their birth
• Birth before or after
survivor parent/s
emigration
• Birth order
• Gender
•
•
•
•
•
•
•
•
•
Age during the Holocaust
Time lapse since the Holocaust
Gender
Type/nature of Holocaust experiences
Reason for persecution
Loss of family
Country of origin
Post-war settlement location
Length of time before resettlement/time spent in
displaced persons camps
• Number of survivor
parents
• Delay between the
end of the war and
their birth
• Birth before or after
survivor parent/s
emigration
• Birth order
• Gender
Grandchildren of Survivor
Generation
• Number of child of
survivor parents
• Birth order
• Gender
Figure 5.2. Addition of Holocaust Survivor Descendant Demographic Moderators to Preliminary Model of the Differential Impact of Holocaust Trauma across Three Generations
© Janine Lurie-Beck 2007
100
Section B
Meta-Analyses of Holocaust Survivor and Descendant Research
© Janine Lurie-Beck 2007
101
Chapter Six – Meta-Analysis Methodology
Chapters Two to Five of this thesis have provided a review of the theoretical and anecdotal
literature with regards to the impacts of the Holocaust on survivors as well as the
transmission of this impact to subsequent generations. In the process of this review, a
Preliminary Model of the Differential Impact of Holocaust Trauma across Three
Generations was developed.
It is necessary to attempt to establish the veracity of this
preliminary model. To this end, it was decided that a detailed and objective assessment of
the existing empirical data be conducted before embarking on a new empirical study. A
meta-analytic review of the empirical assessment of the key variables noted in the
theoretical and anecdotal literature was undertaken. The current chapter explains the
methodological approach to these meta-analyses. Chapters Seven to Ten report on the
findings of this meta-analytic review process.
This chapter provides a detailed explanation of the methodology used for the metaanalytic review of the literature pertaining to Holocaust survivors and their descendants.
Firstly, justifications for the use of meta-analytic techniques as opposed to a traditional
literature review are presented.
Following this, various aspects of the meta-analytic
approach of the thesis are explained. The literature search is described including the
process for identifying relevant research. Various inclusion criteria for analysis are then
described. The nature of the data collected is explained. The meta-analysis calculation
techniques utilised are introduced followed by an explanation of how the results were
interpreted.
The approaches used to counter problems arising from missing data are
clarified, and finally, an explanation of how the results of the analyses conducted have been
presented is provided.
6.1. – Justification for Meta-analytic Methodology
The aim of the first stage of the current thesis was to bring some synthesis to the research
conducted to date with Holocaust survivors, their children and their grandchildren in
relation to demographic variables. A traditional literature review is generally qualitative/
narrative in nature or perhaps utilises a rudimentary quantitative method such as ‘votecounting’ (Egger & Davey Smith, 1997). These methods are open to criticism in relation to
the validity of their conclusions.
‘Vote-counting’ involves producing a simple tally of statistically significant results
and non-statistically significant results.
© Janine Lurie-Beck 2007
Often non-statistically significant results are
102
simply quoted as being non-statistically significant and no information is given as to the
direction of the result. By merely reporting the proportion of statistically significant to
non-statistically significant results the reader of such a review is robbed of vital information
about the relationship in question, such as, an indication of the direction of non-statistically
significant findings and the magnitude of the relationship. A group of individual studies
which may all have non-statistically significant results can lead to the conclusion that there
is a statistically significant difference overall if most of the mean differences are in the
same direction. Such results would lead to the conclusion that no relationship exists in a
‘vote-counting’ exercise (Cook & Leviton, 1980).
The conclusions of such exercises
should clearly be viewed with great scepticism (Cooper, 1984; Egger & Davey Smith,
1997; B. F. Green & Hall, 1984).
Reviews of the same literature can lead to differing conclusions. For example
reviewers may disagree as to which studies to include in their review (Egger & Davey
Smith, 1997). They can also differ in how they choose to weigh the importance of each
study result since this is done in a subjective manner by the reviewer (Egger & Davey
Smith, 1997). Vote-counting usually ignores sample size and so the relatively unbiased
method of weighting the results based on sample size is not utilised (Egger & Davey Smith,
1997).
It has been argued that the same scientific rigor be applied to research literature
reviews as to the individual empirically designed studies. In other words reviews should be
“more technical and statistical than narrative” (Glass, McGaw, & Smith, 1981).
Appropriately applied statistics will enhance the validity of review conclusions (Cooper,
1984). In an interview with Hoffert (1997), Harris Cooper commented that “meta-analysis
gives greater credence to the review because it is an application of a scientific method to a
formerly subjective pursuit.” Hoffert (1997) defines meta-analysis as “a statistical method
of quantitatively combining and synthesizing results from individual studies.”
Meta-analytic reviews have the advantage of taking the magnitude and direction of
each result into account irrespective of their statistical significance level, thus making up
for the shortfall in this area by more traditional review methods.
They allow more
objective appraisal of a body of research which eliminates the problem of differing
opinions when more subjective methods are used (Egger & Davey Smith, 1997).
In addition, meta-analytic reviewers can test hypotheses that were never tested in
individual studies, and, by so doing, add new insight to the area of research. For example
© Janine Lurie-Beck 2007
103
differences between sub-groups of the population or other sources of heterogeneity in
results can be explored via meta-analysis (Egger & Davey Smith, 1997). Thus, metaanalysis should therefore, to some extent at least, reduce the need for new research (B. F.
Green & Hall, 1984).
Meta-analysis involves treating the results of individual studies as a data set for
analysis, the findings of an original study replace the individual as the unit of analysis
(Egger & Davey Smith, 1997; Glass et al., 1981). If thought of in this way, seemingly
contradictory study results can be reconciled very easily. Each study can be viewed as
deriving an estimate of the true population effect by obtaining a result based on a sample
from this population. It is therefore reasonable to assume that study results may differ
purely as a result of sampling from different parts of the population distribution (Lyon, n.d.;
Taveggia, 1974).
Taveggia (1974) goes so far as to say that individual study results are
therefore meaningless because they may have occurred by chance.
meaningful once observed within a distribution of findings.
They are only
By accumulating results
across studies, one can gain a more accurate representation of the population relationship
than is provided by individual studies (Lyon, n.d.).
Numerous reviews have been conducted with regard to impacts of the Holocaust on
the survivors themselves and their children and the interaction between these generations.
An up to date review is required to enable inclusion of more recent research that has not
been incorporated in earlier reviews. The main problem with the majority of reviews to
date is that they have been narrative reviews rather than meta-analytic reviews (for example
Felsen, 1998; Kellerman, 2001c; Solkoff, 1992a). A number of researchers have suggested
meta-analysis of research relating to Holocaust survivors and their descendants (for
example J.J. Sigal & Weinfeld, 1989) however only one meta-analysis was located by the
current author.
Van Ijzendoorn et al. (2003) conducted a meta-analysis relating to
psychological well-being among children of survivors. This meta-analysis only addresses
children of survivors and not three generations as is conducted for the current thesis,
however, several demographic variables of interest were examined. For example, the
impact of having one versus two survivor parents is addressed as well as post-war
settlement location (van Ijzendoorn et al., 2003). However there is one major fault with
their methodology that casts a shadow over their findings.
The research team state that unpublished studies and studies only reported at
meetings or conferences were excluded from the meta-analysis (van Ijzendoorn et al.,
© Janine Lurie-Beck 2007
104
2003). No explanation as to why this is a justified course of action is provided which is of
great concern considering the now established and much discussed bias towards the
publication of statistically significant results (Cook & Leviton, 1980; Cooper, 1984; Egger
& Davey Smith, 1998; B. F. Green & Hall, 1984; Hedges & Olkin, 1985; Hoffert, 1997;
Wolf, 1986). The fact that this approach was used has negative implications for the validity
of their meta-analysis results.
Even among the narrative reviews only one review article
that incorporated unpublished theses was located (Felsen, 1998), however, this review only
included studies conducted in America.
The superiority of meta-analytic review methods over traditional narrative or votecounting reviews has been established. When unpublished studies are also included to
counter publication bias, meta-analysis provides the most objective means of reviewing
research on a given topic and has the added bonus of providing a quantified overall
measure of the relationships assessed in the research.
Therefore, meta-analysis has been
used in the current thesis to examine the empirical support for theories relating to
demographic differentials in the psychological adjustment of Holocaust survivors and their
descendants. The remainder of this chapter details the statistical methodologies used to
conduct the meta-analyses presented in the current thesis.
6.2. – Literature Search Methodology
6.2.1. – Citation Sources
A literature search was performed in order to identify studies conducted with Holocaust
survivors, children of Holocaust survivors and grandchildren of Holocaust survivors
between 1945 and 2006. The search term “Holocaust Survivors” was entered into the
PsycINFO, ProQuest Psychology Journals and Psychology Journals online database
citation databases. Separate searches were conducted with the search term “Concentration
Camp” to check for any articles not using the word Holocaust but studying the same
populations. Translations of works published in languages other than English were sought
and mostly found.
As was discussed earlier, in an exercise such as this, it is vital to include both
published and unpublished research to control, as much as possible, for the much discussed
publication bias towards statistically significant findings (Cook & Leviton, 1980; Cooper,
1984; Egger & Davey Smith, 1998; B. F. Green & Hall, 1984; Hedges & Olkin, 1985;
Hoffert, 1997; Wolf, 1986).
PsycINFO incorporates the Dissertation Abstracts
International database so that both published and unpublished sources were perused. An
© Janine Lurie-Beck 2007
105
additional search was conducted with the WorldCat Dissertations and Theses service (via
OCLC FirstSearch) to ensure adequate search coverage of unpublished research from
around the world.
In addition to computer database searches, search results were cross-referenced with
the comprehensive Holocaust research bibliography, established in 1979 by the late Leo
Eitinger (who conducted some of the earliest research on Holocaust survivors) (Rieck, n.d.a). The most recent print edition produced by Krell and Sherman (1997) was perused.
Miriam Rieck – who helped compile the previous incarnation of the bibliography including
research up to 1985 (Eitinger et al., 1985) – maintains an electronic version of the
bibliography
which
can
be
found
on
the
internet
(http://research-
faculty.haifa.ac.il/arch/index.asp) and this resource was used to check post-1997 references
added to the bibliography (last checked in December 2006).
6.2.2. – Process for Identifying Relevant Articles from Search Results
An initial perusal of books and review articles on the effects of the Holocaust and the
transmission of trauma to subsequent generations identified variables of interest for the
study. These variables were discussed in detail in Chapter Two and were incorporated into
the preliminary Model of the Differential Impact of Holocaust Trauma across Three
Generations. To refresh the reader’s mind, the variables of interest were psychopathology
(specifically anxiety, depression, paranoia and post-traumatic stress disorder symptoms),
post-traumatic growth, world assumptions, coping styles, and romantic adult attachment.
These are labelled as psychological impact variables and influential psychological process
variables in the preliminary versions of the model presented. The search results obtained
for the more detailed meta-analytic investigations were searched for measurements of these
variables.
The results of searches were reviewed via their title and/or abstract. If the title of
the article mentioned a variable of interest it was ear-marked for inclusion. If a title of an
article was too general as to reveal the particular variables assessed in the article, the
abstract was then scanned. Articles with mention of the variables of interest in the abstract
were also then ear-marked. If, after examination of both the title and abstract, it was still
unclear as to whether the particular article was of relevance, the article itself was viewed,
where possible, in an effort to ensure no relevant article was overlooked.
In some instances, a variable of interest was not mentioned specifically but the
study encompassed a broader concept which could encapsulate a variable of interest.
© Janine Lurie-Beck 2007
In
106
particular this refers to articles utilising personality or psychopathology questionnaires
which may include subscales assessing variables of interest. Examples of such measures
include the Minnesota Multiphasic Personality Inventory (MMPI) and the Symptom
Checklist (SCL-90 or SCL-90-R) both of which have subscales measuring depression,
anxiety and paranoia – all variables of interest. If any such broad measures were mentioned
every effort was made to determine the nature of subscales to check if any assessed
variables of interest. If any such measures were mentioned then the study was checked to
see if sub-scale scores were quoted as well as overall scores.
6.3. – Criteria for Inclusion of Studies in Meta-analysis
A number of factors were considered when deciding which studies to include in metaanalyses. These factors include criteria used by researchers when deriving their samples,
the operationalisation of variables measured and methodological quality of studies.
6.3.1. – Criteria used for Sample Selection
The criteria for inclusion in Holocaust survivor, children of Holocaust survivor and
grandchildren of Holocaust survivor groups varies from study to study. The following
broad and inclusive criteria were used in the meta-analyses presented in the current thesis:
• A Holocaust survivor was defined as any person domiciled in a country occupied by the
Nazis and who suffered some form of persecution. This definition is not confined to
those who were interned in concentration camps. It includes people who survived the
war in hiding or in some other way. While many researchers have considered only camp
survivors to be within this category (Felsen, 1998; Hodgkins & Douglass, 1984), the
inclusion of survivors with other experiences allows the analysis of differences between
these groups.
• A child of a Holocaust survivor had to have at least one parent who met the above
criteria for a Holocaust survivor. They could have either one or two Holocaust survivor
parents. They had to be born after the cessation of hostilities. A few studies were found
which included participants born during the war (for example Gertz, 1986): this author
agrees with the argument that these people should be considered survivors themselves
rather than children of survivors (Kestenberg & Kestenberg, 1990a). Sigal and Weinfeld
(1987) duly note that any effects noted in these children could have originated from their
own exposure to Holocaust trauma as opposed to influences from their parents.
Therefore studies of children of survivors including any born pre-1945 were not
included in meta-analytic calculations, but are discussed in the text.
© Janine Lurie-Beck 2007
107
• A grandchild of a Holocaust survivor had to have at least one grandparent meeting the
criteria outlined above for Holocaust survivor and also a parent meeting the criteria for
child of a Holocaust survivor as outlined in the second dot point above, but no parent
meeting the criteria for a survivor (as outlined in the first dot point above).
6.3.2. – Operationalisation of Variables
Only studies including specific operationalisation of a variable being meta-analysed were
included.
For example, the psychopathological variables of depression, anxiety and
paranoia have been subject to such analysis. Only studies including specific depression,
anxiety and/or paranoia measures were included in these analyses. Studies using general
psychopathology/well-being measures or only quoting overall scores for measures
including subscales measuring these variables were not included as meta-analyses were
only conducted for the specific symptoms identified.
6.4. – Collection of Relevant Data/information from Individual Studies
Means, standard deviations, percentages and/or correlation co-efficients as well as sample
sizes were obtained from the articles and dissertations along with the results of significance
testing where conducted. In addition to this statistical information, detailed demographic
characteristics of samples (for example, mean age, age range, gender breakdown, nature of
experiences, religious affiliation), methods utilised to derive the samples and the country
where the study was conducted, were also noted (where reported) so that their influence
could be assessed.
The above described data and information from each study was entered into
Microsoft Excel spreadsheets set up by the current author to conduct significance testing
and the meta-analytic calculations. Significance tests were automatically conducted for all
data entered into the spreadsheet and thus served as a calculation check when such tests had
already been conducted by the original researcher/s. Calculation checks of the original
researcher/s’ significance testing is a practice advocated by Cooper (1984) as a preliminary
process to meta-analysis. All significance tests quoted in the body of the current thesis are
derived from these calculations as opposed to those quoted by the original authors. This
ensures a degree of consistency as the same formula is used to conduct all tests.
© Janine Lurie-Beck 2007
108
6.4.1. – Mean Differences between Two Groups
The following formulae for independent groups t-tests as detailed by Tilley’s (1999)
undergraduate statistics text was used to conduct significance tests for between group mean
differences for each individual study/result:
X1 − X 2
∑x
where σ~ 2 =
t=
N −1
σ~12 σ~ 22
+
N1 N 2
2
Meta-analytic texts consulted (Cook & Leviton, 1980; Cooper, 1984; B. F. Green &
Hall, 1984; Hunter & Schmidt, 1990, 2004; Schwarzer, 1996; Wolf, 1986) suggest the
calculation of an effect size for each group difference be incorporated in a meta-analysis.
An effect size reflects the magnitude or strength of a relationship which a significance test
cannot do with its significant/not significant dichotomy (American Psychological
Association, 2001).
These effect sizes are used for meta-analysis, rather than ts, because they give a
reflection of effect size unfettered by the influence of sample size.
That is, when
calculating a t, the larger the sample size, the larger the t will be with the same original
difference between means. While it is necessary to temper a t-test result with the sample
size when conducting one study, it is unnecessary in a meta-analysis. This is because the
meta-analytic process weights each result in a way that reflects its precision as an estimate
of the population effect.
In addition to this mathematical argument, Hunter and Schmidt
(2004) argue that because the effect size is unaltered by sample size, it is a better reflection
of the population effect size which is defined without reference to sample size.
The effect size, Hedges g, is calculated by dividing the difference between the two
group means by the pooled standard deviation or within-group standard deviation (that is
the weighted [by the degrees of freedom] mean of the two variances square rooted)
(Riopelle, 2000). In formulaic notation:
X1 − X 2
( SD12 × df 1 ) + ( SD 22 × df 2 )
where SD pooled =
g=
SD pooled
(df 1 + df 2 )
Hedge’s g is a positively biased estimate of the population effect size ( δˆ ) when
calculated with small samples. To counter this, Hedges and Olkin (1985, p. 81) provided
© Janine Lurie-Beck 2007
109
the following adjustment formula which was utilised for the meta-analyses reported in the
current thesis:
δˆ = (1 −
3
)g
4N − 9
6.4.2. – Incidence Differences between Two Groups
Studies reporting percentages or incidence levels often did not test for statistical
significance. After an examination of a number of statistical texts it was decided that these
would be tested with the use of odds ratio calculations. The odds ratio was viewed as more
applicable to the analyses required for this project than chi-square, because it is not unduly
affected by uneven sample sizes (Smithson, 2000). In addition, it has become identified
with assessing the significance of differences in incidence levels of symptoms in two
populations (Smithson, 2000) as would be set out in a 2x2 contingency table such as Table
6.1.
Table 6.1.
Example 2x2 contingency table for comparing incidence levels
(X) Group
Survivor
Control
Yes
N11
N21
(Y) Symptom
No
N12
N22
The odds ratio of having the symptom, if in the survivor group, versus the control
group is:
Ω=
Odds (Y = 1 X = 1)
Odds (Y = 1 X = 2 )
where Odds (Y = 1 X = 1) =
N 11
N
and Odds (Y = 1 X = 2 ) = 21
N 12
N 22
N /N
In other words Ω = 11 12
N 21 / N 22
6.5. – Checks for Duplication of Results
Upon examination it became obvious that findings from studies have often been published
more than once. For example the results of a study conducted by Moshe Almagor and
Gloria Leon in 1978/1979 were published in a journal article in 1981 and then in a book
chapter in 1989 (Almagor & Leon, 1989; Leon, Butcher, Kleinman, Goldberg, & Almagor,
1981). In another example, Sophie Venaki, Arie Nadler and Hadas Gershoni’s findings
were published in the International Journal of Social Psychiatry in 1983 and then in
Family Process in 1985 (Venaki, Nadler, & Gershoni, 1983, 1985).
© Janine Lurie-Beck 2007
In yet another
110
example, the results presented in Cohen, Brom and Dasberg (2001) are re-reported with
additional data in Brom, Durst and Aghassy (2002).
Egger and Davey Smith (1998) identify this “multiple publication bias” as of great
concern for meta-analysts. They note that multiple publication bias can further compound
the problem of publication bias because not only are statistically significant results more
likely to be published in the first place, they are also more likely to be published more than
once (Egger & Davey Smith, 1998). If the duplication of data is not recognised, this can
lead to an over-estimation of the meta-analytic effect size. Therefore, special care was
taken to cross-reference study results with at least one common author to ensure (as much
as possible) that duplication of data did not occur in the current project.
The issue of possible duplication of data also arose when looking at dissertations
and theses which were later published as journal articles or book chapters, as was the case
for the work of Helen Lichtman, Susan Rose and Felice Zilberfein (Lichtman, 1983, 1984;
Rose, 1983; Rose & Garske, 1987; Zilberfein, 1994, 1996).
All dissertations and theses
obtained were cross-referenced with the published material to ensure data was not
duplicated. Dissertations are a useful resource, as they contain more detailed presentation
of results than journal articles do and so were often consulted in the search for missing data.
6.6. – Calculation Methods for Meta-Analyses
Meta-analyses were conducted on two types of statistics. These are the examination of
mean differences between two groups as would be analysed by an independent groups t-test
and the examination of differences in incidence levels between two groups as would be
analysed by an odds ratio. It should be noted that the calculation methods used were crosschecked by running representative meta-analyses on a number of recognised meta-analytic
computer programs such as Comprehensive Meta-Analysis, Meta-Stat and Stats Direct.
Excel calculation spreadsheets were developed by the author and used for the metaanalyses in this thesis, in preference to the use of meta-analytic computer programs, to be
certain that the inner calculations were being conducted correctly. Also not all of the
analysis tools required for the desired analyses for this thesis were available in any given
program located.
© Janine Lurie-Beck 2007
111
6.6.1. – Mean Differences between Two Groups
Meta-analytic or weighted average effect sizes were derived by weighting each effect size
by the inverse of its variance as advocated by Hedges and Olkin (1985). The individual
effect size variance was calculated with the following Hedges and Olkin (1985) formula:
 δˆ 2
SDg2 = 
 2df
  N 
+

 n n 
  1 2
The meta-analytic effect size was then calculated as follows:
1
Σ(δˆ ×
)
SDg2
g=
1
Σ( 2 )
SDg
6.6.2. – Incidence Differences between Two Groups
Meta-analytic odds ratios were calculated by simply producing a sum of all frequencies in
each cell for all the relevant individual odds ratios (the formula for which is presented in
Section 6.4.2) and then calculating an overall meta-analytic odds ratio based on the total
population of people included in all studies.
In other words,
∑N
11
,
∑N
12
,
∑N
21
and
∑N
22
were derived and then these
figures were entered into the odds ratio formula as follows:
Ω=
∑N /∑N
∑N /∑N
11
12
21
22
6.7. – Sub-Group Meta-Analyses
In addition to overall meta-analyses, meta-analyses of subgroups were also conducted
where viable. Sub-groups were either location groups or other demographic variables such
as gender.
As has already been discussed, a large majority of studies have treated Holocaust
survivors, children of Holocaust survivors and grandchildren of Holocaust survivors as
homogenous groups and have compared their entire sample’s combined scores to control
groups. A lot of these studies, however, quote data relating to subgroups within their study
groups (for example males versus female groups, children of one versus two survivors).
While some have checked for statistically significant differences between these subgroups,
© Janine Lurie-Beck 2007
112
others have not conducted significance testing at all, citing means and standard deviations,
often in an appendix. Consequently there were many opportunities to assess differences
between subgroups that the original researchers left unrealised or perhaps just unreported.
As was discussed in Chapter Four, markedly different post-Holocaust experiences
may have been encountered by survivors depending on where they chose to settle after the
war. Three distinct locations/location groups were considered for possible meta-analysis.
These fall roughly into one of three post-war settlement groups of those who stayed in
Europe, those who went to Israel and those who immigrated to other continents such as
America. It is possible to examine this issue via sub-group meta-analyses by grouping
studies based on the country or region they were conducted in.
One of the main aims of this research project was to examine demographic
differences within groups of survivors, children of survivors and grandchildren of
survivors. Therefore, where sufficient results were obtained, separate meta-analyses were
conducted to assess such differences. These most commonly related to gender and one
versus two survivor parents for children of survivors, although in some instances other
demographic variables could be meta-analysed.
However, it is noteworthy that such
analysis could only be conducted if the original author quoted data for these subgroups, or
their entire sample was from the subgroup. Davey Smith and Egger (1998) point out that
the decision to include sub-group data may have been dependent on the statistical
significance of the results. This should be borne in mind when interpreting these results.
6.8. – Meta-Regression: Correlation of Study Effect Sizes with Study Attributes
Meta-regression was also used to examine the relationship between the strength of an effect
and variables such as age and time lapse since the Holocaust. This involved correlation of
effect sizes with the demographic characteristics of age, gender (for example female
percentage of the sample), number of Holocaust survivor parents (for example percentage
of the sample with two survivor parents), time between the study and the Holocaust, and
time between the Holocaust and the birth of children of Holocaust survivors. Davey Smith,
Egger and Phillips (1997) describe how meta-regression can be used to look at the
“gradient in .. effects”. They state that:
Such a gradient allows for a more powerful examination of differences in
outcomes, as a statistical test for trend can be performed.… Attributes of
study groups such as age and length of follow-up can readily be analysed
in this way.
© Janine Lurie-Beck 2007
113
Meta-regression analyses were conducted using SPSS.
The effect size and
demographic data used from each study were weighted by the survivor or descendant
sample size, so that studies with larger sample sizes were given more weight in the
analysis.
The formula followed by SPSS for this calculation is equivalent to the formula for
Pearson’s correlation co-efficient as presented in Tilley (1999) with the added element of
weighting by sample sizes:
rmeta =
∑ (( Z Z ) N
∑N
X
Y
HGroup
)
Hgroup
Scatterplots were viewed before each analysis was conducted to ensure there was no
evidence of a curvilinear (as opposed to linear) relationship and that Pearson’s correlation
co-efficient was an appropriate statistic to apply to the data.
Not all researchers reported the data required for these meta-regressions (for
example the average age of their subjects or the gender breakdown) which means that not
all studies could be included in meta-regressions. This factor must be kept in mind when
interpreting the results of the meta-regression. Where this occurs, a note is made informing
the reader so that the results can be interpreted in an informed manner.
6.9. – Criteria for Inclusion of Multiple Results from Single Studies in Meta-Analysis
Hunter and Schmidt (1990; 2004) identify several conditions where it is statistically valid
to include multiple results from one study in meta-analytic calculations. These are when
the study essentially incorporates a “fully replicated design” or when “conceptual
replication” occurs.
A fully replicated design occurs when the same measure or condition is used with
independent groups.
Examples relevant to the current research would be measures
conducted with two groups of Holocaust survivors or other generations sourced from two
different locations or different sample sources such as survivor groups versus clinical
settings, or with demographic subgroups such as males and females or differing age groups.
The key determinant is that the groups are independent and that no participant is included
in both groups.
Conceptual replication is present when there is replicated measurement of a
particular variable. For example, Finer-Greenberg (1987) uses both the Symptom Check
List-90-Revised (SCL-90-R) and the Cognitive Checklist to measure depression and
© Janine Lurie-Beck 2007
114
anxiety levels in Holocaust survivors. This means that for each Holocaust survivor there
are two scores for depression and two scores for anxiety.
Including each separate
conceptual replication violates the assumption of independence (Hunter & Schmidt, 1990,
2004). This assumption states that each result included in the meta-analysis must be
independent from the other results included.
Therefore, effect sizes representing
conceptual replications were averaged so that the independence assumption was not
breached. This also ensured that each study participant contributed to the meta-analytic
result once.
6.10. – Methods for Dealing with Missing Data
A common problem faced by meta-analysts is missing/unreported data or statistics (Cooper,
1984). Thankfully, there were very few instances of missing data in the research included in
the meta-analyses for this project. For the rare instances where these problems occurred an
effort was made to calculate a replacement value where the missing data would impact on
the meta-analytic calculations. Wherever this practice was conducted a note is made in the
text so that the reader is aware and can interpret analysis results in the light of this. The
following sections outline the types of missing information that were confronted and how
they were dealt with.
6.10.1. – Missing Standard Deviations
A weighted average of the standard deviations of all groups in all other comparisons using
the same questionnaire was used as a replacement standard deviation. This was obtained
by calculating the average weighted variance and then square rooting this figure to derive
the weighted average standard deviation. It was not necessary to calculate a replacement
standard deviation, if the results of significance testing were reported. A replacement value
was only calculated, if it was required to derive an effect size for the meta-analysis.
6.10.2. – Presentation of Significance Tests without Means or Standard Deviations
A formula quoted by several authors has been developed in order to calculate a g-value, for
use in a meta-analysis, when only a t-value is reported without the means and standard
deviations of the groups involved. The following formula (Cooper, 1984; Lyon, n.d.) was
utilised to convert t-values to g-values when this circumstance arose:
g=
2t
df
© Janine Lurie-Beck 2007
115
When only a p-value was quoted, the TINV function in Microsoft Excel was used as
it returns the t-value when given the p-value and the degrees of freedom for a particular
significance test. The method stated in the previous paragraph was then used to derive an
effect size from this derived t-value.
6.10.3. – Statement of Result without Data or Statistics Reported
Sometimes authors merely cited whether they obtained a statistically significant result
without reporting any descriptive statistics at all. In these instances, the most unbiased
estimate of a t-value was inserted into the analysis.
The most problematic situation occurred where an author obtained a nonstatistically significant result. It is not uncommon in such circumstances for no data to be
reported and a sentence such as “There was no statistically significant difference between
the survivor group and the control group on the XXX measure” included in the text.
However, it is obviously important for the integrity of the meta-analytic process for null
results to be included as well as statistically significant results, so that an accurate overall
effect is calculated. Unless this is stated in the article, it is also often impossible to know in
which direction the non-statistically significant result was obtained.
Because of the difficulties just described, the most commonly suggested solution is
to assume a t or effect size of zero (Cooper, 1984). Given the lack of information, this is
the only unbiased replacement value that can be inserted. It is important to note, however,
that an effect size of zero is rare and that the non-statistically significant result will have
almost certainly been in a certain direction. Given this, it must be borne in mind that using
zero as a replacement value will have the effect of making the overall effect size, or g ,
closer to zero than it might actually be (Cooper, 1984).
It was less common for a researcher to state that a statistically significant difference
was found between the two study groups without citing any descriptive statistics or test
results. In these instances, the minimum t-value required to derive a statistically significant
result at a significance level of 0.05 for a two-tailed test (given the sample size/degrees of
freedom) was inserted into the analysis.
Again, while this is by no means a perfect
solution, it does not impact upon the analysis as much as missing information regarding a
non-statistically significant result which can vary much more markedly in terms of
direction and degree (that is, between the negative and positive values of the cut-off t -value
required for significance).
© Janine Lurie-Beck 2007
116
6.11. – Interpretation of Meta-Analytic Findings
With weighted overall effect sizes and odds ratios calculated it was necessary to have a
framework for interpreting the meaning of the results. In addition to determining statistical
significance, two supplementary interpretation methods were used. “Fail Safe N” was
calculated to address possible reader concerns relating to studies which may have been
inadvertently left out of the analysis for various reasons. The issue of homogeneity of effect
sizes was also addressed.
6.11.1. – The File Drawer Question
When conducting meta-analysis it is desirable, but often impossible, to include every piece
of research conducted to date on the issue of interest. It is possible that there are a number
of studies that are left uncovered because they remain unpublished or unobtainable or do
not show up in database searches because they have not been added to them or do not come
up with the search terms used. This relates back to the idea that there is a publication bias
towards statistically significant results which implies that there may well be studies with
non-statistically significant results that were rejected for publication because of this. It
should be noted again that for the meta-analyses outlined in the current thesis every effort
was made to include unpublished data as well as published data. However, it needs to be
acknowledged that the chances of obtaining the data from every single unpublished study
ever conducted are very low, despite the most fervent of efforts by a meta-analyst. In light
of this, a formula has been developed which estimates the number of opposite results
needed to produce a non-statistically significant overall effect. This has also been called
the “Fail Safe N” (Cooper, 1984; Hunter & Schmidt, 1990, 2004; Rudner, Glass, Evartt, &
Emery, 2002).
The formulae used for this project are based on that presented by Hunter and
Schmidt (1990; 2004) and Orwin (1983). The maximum meta-analytic effect size, or
g critical required so that its 95% confidence interval would encompass zero (and therefore be
considered non-significant) is determined. The following formula then calculates how
many additional non-statistically significant or opposing results would be required to
reduce g to that level.
k ( g − g critical )
gc
where k0 is the additional number of opposite results required and k is the number of
k0 =
comparisons included in the analysis.
© Janine Lurie-Beck 2007
117
6.11.2. – Testing the Homogeneity of Effect Size Sets
The main premise of conducting a meta-analysis is the hope of creating the most precise
estimate of the true size of a population effect. Related to this premise is the question of
whether it is valid to assume that the collection of effect sizes/study results being
aggregated in a meta-analysis are a homogeneous set, related to a single population effect
size or whether they are actually a heterogeneous set, reflective of a number of different
populations which should be considered separately. A statistical test designed to assess
whether a set of effect sizes is homogeneous is the Q Homogeneity Test. The formula
(Hedges & Olkin, 1985, p. 153; Kline, 2004, p. 260) for this test is:
Q=∑
( g − g )2
where the SDg2 formula is given in Section 6.6.1
2
SDg
The result of this test is assessed via the χ 2 distribution with k – 1 (k = number of
studies/results) degrees of freedom.
The assumption of a single population effect size is implicit in a “fixed-effects”
meta-analytic model.
When this assumption has been breached, as evidenced by a
statistically significant Q Homogeneity Test result, it must be assumed that the set of effect
sizes/study results represent a heterogeneous group. The courses of action available to the
researcher are either to conduct meta-analyses using a “random-effects” model (which does
not assume a single population effect size) or to conduct sub-set meta-analyses (Kline,
2004, p. 260).
One of the main objects of the current thesis is to identify the demographic subgroups of survivors and their descendants that differ statistically significantly from each
other. Therefore, the use of a “random-effects” meta-analysis model is not of relevance
here. It was not the aim of the study to acknowledge but not explore the sources of
heterogeneity. The aim of the meta-analytic review of the literature was to establish
statistically the presence of heterogeneity and then to attempt to uncover the reasons for it.
To test for heterogeneity it is necessary to use a meta-analytic tool designed to test whether
a set of study results are homogeneous in the hopes that the results of such a test would be
negative.
Therefore the Q Homogeneity Test is an important tool in answering the
questions raised by the literature review of survivor research. Beyond establishing the
heterogeneity of results, the methodology required to examine demographic differences is
the use of sub-set meta-analyses.
© Janine Lurie-Beck 2007
118
6.12. – Overview of Meta-Analysis Section of Thesis
This chapter has summarised and explained the meta-analytic methods used in Stage One
of this thesis. The rest of this section contains the results of literature reviews in which
these methods were used. Chapter Seven reports the results of studies which compare
Holocaust survivors, children of survivors and grandchildren of survivors to
control/comparison groups. Chapter Eight presents meta-analytic and non-meta-analytic
reviews of demographic differences within the Holocaust survivor group. Chapter Nine
addresses demographic differences among the descendants of Holocaust survivors, namely
their children and grandchildren. The reader should note that all studies included in the
meta-analyses/review of the literature will be discussed in past tense while the results of
meta-analyses will be discussed in present tense. This will aid the reader in differentiating
between results of studies in the literature and results of the meta-analyses conducted for
the current thesis. In Chapter Ten the model of the differential impact of Holocaust trauma
is revised in light of the meta-analytic and non-meta-analytic review findings. The results
of this process are examined to determine which areas have convincing evidence and which
are more ambiguous and require further study. This review and reappraisal process informs
Stage Two of the thesis which entails an empirical assessment of the revised model.
© Janine Lurie-Beck 2007
119
Chapter Seven – Meta-Analyses of Survivor and Descendant Groups versus Control
Groups/General Population
Opinions have been divided as to the extent Holocaust survivors and their descendants
differ from the general population in terms of psychological health and functioning. An
examination of the literature clearly reveals a range of results with some studies finding
statistically significant differences between survivors (or their descendants) and the general
population (as represented by control groups) and others finding negligible differences or
no differences at all. While arguments such as differing sample sources are put forward as
reasons for the heterogeneity of findings, the point that perhaps results differ so widely
because groups of survivors may differ widely in their level of post-Holocaust adjustment
has rarely if ever been put.
This chapter provides a quantitative synthesis (via meta-analyses) of findings to
date which compare Holocaust survivors or descendant groups to control groups/the
general population on a number of psychological variables. In so doing it aims to establish
the current state of evidence in relation to differences between Holocaust survivors and
descendants and the general population. The specific meta-analysis hypotheses (MAH) are
that:
MAH1:
Survivors, children and grandchildren will, overall, have statistically
significantly higher scores on negative psychological variables and statistically
significantly lower scores on positive psychological variables than control groups/the
general population when the currently available data in the literature is synthesised via
meta-analyses. In other words, overall differences BETWEEN survivor and descendant
groups and control groups will be statistically significant.
MAH2: The individual effect sizes and odds ratios included in the meta-analyses will
have large ranges and the Q test of homogeneity of meta-analytic effect sizes will be
statistically significant. These results will suggest that survivors and descendants differ
statistically significantly from the general population, but the size of this difference
varies substantially from study to study - a finding suggestive of the fact that the
survivor and descendant population is a heterogeneous rather than homogenous group
which warrants WITHIN group/sub-group comparisons.
© Janine Lurie-Beck 2007
120
7.1. – Method
Studies comparing Holocaust survivors or descendants to control groups were collated for
meta-analyses.
The psychological variables addressed by these meta-analyses are
depression, anxiety, paranoia, world assumptions of benevolence and meaningfulness and
positive and negative romantic attachment and intimacy dimensions. Because the range of
measures used to assess attachment and interpersonal trust and intimacy were so varied,
studies/results were grouped under the broad headings of positive or negative dimensions.
The studies included report results based on surveys or questionnaires (where mean
ratings are provided) or on diagnoses or categorisations (where percentages of samples
qualifying for a given diagnosis or category are provided). Separate meta-analyses are
conducted for sets of studies reporting on survey means versus those reporting on
percentages qualifying for diagnoses. Meta-analytic statistics reported include effect sizes,
odds ratios, Fail Safe Ns and Homogeneity Q tests (refer to Chapter Six for detailed
explanations of these analyses).
To be included in the analyses studies had to include a survivor or descendant group
and a control group that were measured at the same time. Because of potential confounds
due to historical and geographical factors the few studies that compared survivor or
descendant groups to normative data or data sourced from a separate study, to serve as a
control group, were not included in the meta-analyses but are discussed in the text. The
studies that included survey ratings of participants by third parties were also not included so
that all analyses only include self-rating on surveys.
7.2. – Holocaust Survivors versus Control Groups
7.2.1. – Meta-analytic Results
In this section the results of studies comparing Holocaust survivors to control groups are
combined meta-analytically and discussed. Table 7.1 presents the results of meta-analyses
of survey studies comparing Holocaust survivors to control groups. Table 7.2 presents the
results of meta-analyses of incidence or diagnosis studies. The individual studies/results
that were included in each of these analyses can be found tabulated in Appendix A.
© Janine Lurie-Beck 2007
121
Table 7.1.
Summary of meta-analyses of survey/scale studies comparing survivors to control/comparison groups
Variable
Depression
Anxiety
Paranoia
Assume World is
Benevolent
Assume World is
Meaningful
Positive
Attachment/Intimacy
Variables
Negative
Attachment/Intimacy
Variables
Number of
results
where
survivors
scored
higher than
comparison
group
29/30
12/13
10/12
3/9
0.32 *
0.57 *
0.45 *
- 0.28 *
3/9
95% confidence
limits for g
Fail
Safe N
Highest
individual
result effect
size (g)
Survivor
group
sample size
0.26 to 0.39
0.47 to 0.67
0.34 to 0.56
- 0.16 to - 0.40
123
62
37
31
1.96
1.41
0.99
-0.50
2,000
907
660
316
1,880
649
645
470
114.41 *
175.53 *
29.43 *
24.64 *
- 0.11
- 0.23 to 0.00
18
-0.53
316
470
25.73 *
0/3
- 0.26 *
- 0.02 to - 0.51
6
-0.64
130
85
4.79
2/3
0.28 *
0.03 to 0.52
<1
0.84
130
85
14.70 *
g
Control
group
sample
size
Q
Homogeneity
test
Note. A positive effect size ( g ) indicates that survivors scored higher on a variable than control groups while a negative
effect size indicates that survivors scored lower on a variable than control groups.
* p < 0.05
Table 7.2.
Summary of meta-analyses of incidence/diagnosis studies comparing survivors to control/comparison groups
Variable
Depression
Anxiety
Paranoia
Insecure
Attachment
Number of
results with
higher
incidence
for survivor
group
5/5
4/4
2/2
1/1
Ω
95%
confidence
limits for
1.27 *
1.44 *
1.24
2.87 *
Ω
1.11 to 1.46
1.27 to 1.63
0.99 to 1.55
1.20 to 6.86
Average
incidence
among
survivors
33%
48%
13%
77%
Survivor
incidence
range
18% to 61%
8% to 64%
12% to 54%
-
Survivor
group
sample
size
Average
incidence
among controls/
comparisons
3,488
3,470
2,353
48
28%
44%
54%
Control
incidence
range
13% to 46%
20% to 50%
9% to 50%
-
Control
group
sample
size
1,561
1,570
1,190
48
Note. The average incidences for depression and anxiety include the findings of Eitinger (1972) and Matussek (1975)
while the average incidence for paranoia also includes the findings of Matussek (1975). These two studies did not include
comparisons to control groups and so could not be included in the odds ratio analysis.
Ω meta-analytic odds ratio
* p < 0.05
Overall, Holocaust survivors score statistically significantly higher on negative
psychological variables and statistically significantly lower on positive psychological
variables than control groups. Meta-analytic effect sizes based on survey studies range
between 0.28 and 0.57 for negative variables and -0.11 and -0.28 for positive variables.
When the mode of measurement is incidence or diagnosis rather than survey scores,
survivors also have a higher incidence of negative psychological symptoms or disorders
than control groups. The odds of survivors being categorised as depressed, anxious or
paranoid are between 1.24 and 1.44 times higher than for members of control groups with
the odds of being categorised with an insecure attachment being 2.87 times higher.
© Janine Lurie-Beck 2007
122
However, there is also a large range in effect sizes and incidence levels. This
suggests that even though, when all the data is collated, survivors differ statistically
significantly from the general population, there is substantial variation within the survivor
population as to the size of that disparity. For example, while survivors included in the
data in the meta-analysis of depression survey results scored an average of 0.32 of a
standard deviation higher than their control group, the most extreme difference between a
survivor group and a control group for depression was 1.96 standard deviations (see Table
7.1). This large variation is further evidenced by the predominantly statistically significant
Homogeneity Q tests which suggest that the set of effect sizes are heterogeneous and do not
represent a homogeneous population.
7.2.2. – Studies Excluded from Meta-analyses
Three studies asked children of survivors to rate their survivor parents on psychological
variables (Major, 1996; Podietz et al., 1984; Woolrich, 2005). Podietz et al. (1984) asked a
group of children of survivors to rate their survivor parents in terms of how fearful they
perceived them to be. The researchers did not cite the descriptive data of the groups but did
cite the statistically significant t test results. Children of survivors rated their mothers and
their fathers as statistically significantly more fearful than control mothers and fathers were
rated (t (216) = 3.63, p < 0.001 and t (216) = 2.58, p < 0.05 respectively). In Major’s
(1996) study, participants rated the extent to which they believed their parents were
suspicious of others. There were no differences in how children of survivors and children
whose parents were not survivors rated their parents on this variable. Survivor mothers
were rated identically to mothers who had not survived the Holocaust (M = 2.50 for both
groups) while survivor fathers were rated only slightly lower than comparison fathers (M =
2.40 versus M = 2.50). No standard deviations were reported but results of t-tests were and
there were no statistically significant differences in ratings here. In the most recent of the
studies to report children of survivor’s ratings of their parents, Woolrich (2005) reported
the percentage of participants who considered their parents general mood to be
characterised by anxiety or depression. Survivor mothers were more likely to viewed as
anxious (29% of 59 versus 19% of 59) than comparison mothers but were equally likely to
be viewed as depressed (27% for both groups). Survivor fathers were more likely to be
viewed as anxious (21% of 69 versus 12% of 59) and depressed (26% versus 12%) than
fathers who had not survived the Holocaust. When analysed via odds ratio, none of these
differences reached significance. The findings of these three studies do not contradict the
© Janine Lurie-Beck 2007
123
overall finding that in general Holocaust survivors score higher or have a higher incidence
of negative psychological dimensions and score lower or have a lower incidence of positive
psychological dimensions.
7.3. – Children of Holocaust Survivor/s versus Control Groups
7.3.1. – Meta-analytic Results
This section summarises and meta-analyses the results of studies comparing children of
survivors to control groups. Table 7.3 summarises the findings of meta-analyses of survey
studies while Table 7.4 presents the findings of odds ratio analysis of incidence studies.
Information regarding the individual studies/results included in these meta-analyses is
presented in Appendix B.
Similar to the findings pertaining to survivors, children of survivors are seen to
overall evidence higher levels/incidence of negative psychological symptoms and lower
levels of positive psychological dimensions.
The effect sizes are smaller than those
obtained for the survivor analyses suggesting less disparity with the general population one
generation removed from the trauma.
While there is a sizeable range of effect sizes in the survey study meta-analyses,
none of the Q homogeneity tests are statistically significant. However it is notable that a
number of them are very close to the significance threshold.
This may suggest less
heterogeneity in the children of survivor population than the survivor population.
Table 7.3.
Summary of meta-analyses of survey/scale studies comparing children of survivors to control/comparison groups
Variable
Depression
Anxiety
Paranoia
Assume World is
Benevolent
Assume World is
Meaningful
Positive
Attachment/Intimacy
Variables
Negative
Attachment/Intimacy
Variables
Number of results
where children of
survivors scored
higher than
comparison group
17/26
13/17
7/12
1/1
g
0.10 *
0.18 *
0.21 *
0.00
95%
confidence
limits for g
0.01 to 0.19
0.06 to 0.29
0.04 to 0.39
- 0.33 to 0.34
Fail
Safe
N
Highest
individual
result effect
size (g)
Children of
survivor
group
sample size
Control
group
sample
size
Q
Homogeneity
test
3
9
3
-
0.56
1.02
0.78
-
1,087
613
298
67
770
482
164
70
28.06
19.52
19.93
-
-
-
67
70
-
0/1
-
3/8
- 0.09
- 0.22 to 0.04
-
- 0.43
418
379
14.57
6/13
0.02
- 0.08 to 0.13
-
0.35
658
510
15.78
Note. A positive effect size ( g ) indicates that children of survivors scored higher on a variable than control groups while
a negative effect size indicates that children of survivors scored lower on a variable than control groups.
* p < 0.05
© Janine Lurie-Beck 2007
124
Table 7.4.
Summary of meta-analyses of incidence/diagnosis studies comparing children of survivors to control/comparison groups
Variable
Depression
Anxiety
Insecure
Attachment
Number of
results with
higher
incidence
for child of
survivor
group
3/4
4/4
1/1
Ω
6.70 *
3.14 *
2.99 *
95% confidence
limits for
Ω
4.24 to 10.58
2.08 to 4.76
1.23 to 7.25
Average
incidence
among
children of
survivors
Children of
survivor
incidence
range
Child of
survivor
group
sample
size
Average
incidence
among controls/
comparisons
Control
incidence
range
25%
24%
68%
3% to 56%
7% to 70%
-
399
399
48
5%
7%
42%
2% to 12%
5% to 10%
-
Control
group
sample
size
580
580
50
Note. Both average incidences also include the findings of Zajde (1998). Zajde (1998) did not compare her results to a
control group and so her data could not be included in the odds ratio analysis.
* p < 0.05
7.3.2. – Studies Excluded from Meta-analyses
Five studies were omitted from the meta-analyses comparing children of survivors to
control groups because of design flaws. Three of these included participants in their
children of survivor group that do not meet the criteria for “child of survivor” as outlined in
the meta-analysis methodology chapter of the current thesis (specifically birth post-1945).
The remaining two studies had problems surrounding their “control” groups.
Stein’s (1997) child of survivor group included participants who were born as far
back as 1922. Clearly such participants could be classified as survivors themselves, despite
technically also being the children of Holocaust survivors. Because the definition of a child
of survivors for the current thesis is a person who was born after their parents’ persecution
had ended and who had not directly experienced the Holocaust themselves this sample was
considered contaminated. However, it is worthy to note that Stein’s (1997) results provide
further support for the trend of higher scores among children of survivors on negative
psychological variables. Stein (1997) conducted a study assessing anger expression before
and after a mood-inducing condition. He also assessed sadness levels before and after the
induction. The sadness levels of children of survivors and the control group prior to the
induction can be seen as another measure of depression in this group.
Scores were
averaged over the induction groups to provide total group scores for the child of survivor
and the control group. It was the children of survivors who were statistically significantly
sadder than the control group (M = 50.27, SD = 12.49, n = 52 versus M = 44.12, SD = 4.88,
n = 51). This difference was statistically significant (t (101) = 3.27, p < 0.01).
Gertz’s (1986) study comparing 111 children of survivors to 53 controls could not
be included in the meta-analysis because of the sample criteria used. Subjects aged up to
43 at the time of the study were included despite the fact that they must have been born no
© Janine Lurie-Beck 2007
125
later than 1943: clearly before the end of the war. The daughters of survivors scored higher
on the Beck Depression Inventory than female controls (M = 8.28, SD = 6.39, n = 60 versus
M = 6.65, SD = 5.67, n =32). A similar pattern of results was noted for the males (sons of
survivors M = 7.82, SD = 9.35, n =51 versus controls M = 5.04, SD = 4.55, n = 21). On the
Spielberger Trait Anxiety Scale both child of survivor groups also scored higher than their
respective control groups (Female M = 42.35, SD = 10.29 versus M = 40.43, SD = 9.95;
Male M = 40.34, SD = 12.35 versus M = 38.70, SD = 9.79). None of these differences
reached statistical significance but were all in the same direction.
In the third study excluded because of sample contamination, Rubenstein (1981)
examined differences in depression between children of survivors and a control group.
This study also included participants born before 1945 and was thus conducted with a
contaminated sample. Despite the problems with Rubenstein’s (1981) study, his findings do
add further support to the idea that children of survivors show higher levels or greater
severity of psychological symptoms than control groups or the general population.
Rubenstein (1981) divided his child of survivor sample into children with one survivor
parent and children with two survivor parents. Both child of survivor groups scored
statistically significantly higher on the Depression scale of the Mini-Mult (an abbreviated
version of the MMPI) than the control group (children with one survivor parent M = 20.75,
SD = 4.52, n = 48, children with two survivor parents M = 22.40, SD = 4.88, n = 30, control
M = 18.41, SD = 3.02, n = 24; t (70) = 2.57, p < 0.05; t (52) = 3.62, p < 0.001).
Utilising the SCL-90-R, Chayes (1987) compared depression, anxiety, phobic
anxiety and paranoia levels of a sample of children of survivors to non-patient normative
data. Normative data is not the same as a contemporaneous and geographically proximate
control group. In addition, the large sample used to derive the normative data (n = 974)
was so large that it was having a significant impact on the meta-analytic results, and so the
study was left out of the analysis. Chayes’ (1987) children of survivors had lower levels of
depression (M = 0.28, SD = 0.56, n = 25 versus M = 0.36, SD = 0.44, n = 974), paranoia (M
= 0.24, SD = 0.64; M = 0.34, SD = 0.44) and anxiety (M = 0.21, SD = 0.51; M = 0.30, SD =
0.37) than the normative data. The children of survivors scored higher than the normative
data on phobic anxiety (M = 0.17, SD = 0.61; M = 0.13, SD = 0.31). None of these
differences reached statistical significance however.
Finally, Berger (2003) compared the results of her sample of 109 children of
survivors to a sample of college students (derived from another study) on the Experiences
© Janine Lurie-Beck 2007
126
of Close Relationships Scale which yields scores on the attachment scales of avoidance and
anxiety. No statistically significant differences were found but the children of survivors
scored slightly higher on both scales (Avoidance M = 49.43, SD = 23.29 versus M = 47.63,
SD = 18.59; Anxiety M = 64.72, SD = 25.92 versus M = 63.06, SD = 21.26).
7.4. – Grandchildren of Holocaust survivors versus Control Groups
7.4.1. – Meta-analytic Results
This section summarises the meta-analyses of results comparing grandchildren of survivor
to control groups.
There were no studies looking at the incidence of symptoms or
diagnoses so the results here only pertain to survey studies and are presented in Table 7.5.
Table 7.5.
Summary of meta-analyses of survey/scale studies comparing grandchildren of survivors to control/comparison groups
Variable
Depression
Anxiety
Paranoia
Positive
Attachment/Intimacy
Variables
Negative
Attachment/Intimacy
Variables
Number of results
where grandchildren
of survivors scored
higher than
comparison group
1/2
3/4
0/1
0/3
g
95%
confidence
limits for g
Fail
Safe
N
Highest
individual
result effect
size (g)
Grandchild of
survivor group
sample size
Control
group
sample
size
Q
Homogeneity
test
0.41 *
0.43 *
- 0.23
- 0.43 *
0.07 to 0.76
0.15 to 0.72
- 0.74 to 0.27
- 0.81 to 0.04
<1
2
5
0.84
0.96
- 0.49
70
109
30
53
64
76
30
45
6.68 *
7.99
0.32
2/2
0.46
- 0.22 to 1.13
-
0.84
11
15
0.84
Note. A positive effect size ( g ) indicates that grandchildren of survivors scored higher on a variable than control groups
while a negative effect size indicates that grandchildren of survivors scored lower on a variable than control groups.
* p < 0.05
As was the case for the survivors and the children of survivors, grandchildren of
survivors score statistically significantly higher on negative variables and statistically
significantly lower on positive variables than control groups. Interestingly the effect sizes
found here are higher than those found for the children of survivors and are not consistent
with the suggestion of a dissipation of effect with each generational removal. However it is
not clear if this is a genuine effect or perhaps due to the small number of studies and
grandchildren of survivors included in these analyses. Certainly the findings for survivors
and children of survivors can be considered more robust because of the larger sample sizes
they are based on.
7.4.2. – Studies Excluded from Meta-analyses
Two studies presented ratings by third parties of grandchildren of survivor versus control
groups and were therefore excluded from the meta-analyses. Rubenstein (1981) reported
teacher ratings of primary school aged grandchildren of survivors while Sigal and Weinfeld
© Janine Lurie-Beck 2007
127
(1989) asked parents and teachers to rate their grandchildren of survivor sample. Sigal and
Weinfeld (1989) had two groups of grandchildren of survivors: one group with at least one
Holocaust survivor grandparent (n = 58) and one group with at least one survivor parent
and one survivor grandparent (n = 11). Two control groups were also used: one group with
native born grandparents (n = 30) and one group with at least one pre-war immigrant
grandparent (n = 28). The percentages of parents and teachers who considered the children
were sad often or very often were:
23% of grandchildren of native born, 30% of
grandchildren of pre-war immigrants, 21% of grandchildren of survivors and 57% of
children/grandchildren of survivors. While the sample size was very small it is interesting
to note that the subjects with a parent and a grandparent who survived the Holocaust were
rated as sad often or very often much more frequently than the other groups.
The
percentages for often or very often fearful or anxious were: 11% of grandchildren of native
born, 19% of grandchildren of pre-war immigrants, 34% of grandchildren of survivors and
40% of children/grandchildren of survivors.
In addition to self-ratings which were
included in the meta-analysis, the grandchildren in Rubenstein’s (1981) study were rated on
anxiety on the School Behaviour Checklist by their school teachers. Children with survivor
grandparents scored higher than the control group but not statistically significantly so (two
survivor grandparents M = 51.92, SD = 10.64, n = 15; one survivor grandparent M = 46.72,
SD = 9.03, n = 24; control M = 45.09, SD = 7.62, n = 11). These findings further add to
overall finding that grandchildren of survivors evidence higher symptom levels than the
general population.
7.5. – Summary and Conclusions
This chapter has presented the findings of meta-analyses comparing Holocaust survivor and
descendant groups to control groups. The aims of this exercise were: 1) to establish the
extent to which survivors and their descendants differ to the general population (as
represented by control groups) on a number of psychological variables and; 2) to determine
if the size of this difference between survivor and descendant groups and the general
population is homogeneous (suggesting that survivors and their descendants are a
homogenous group).
The summation of results from studies comparing survivor and descendant groups
to control groups does support the argument that OVERALL survivors and their
descendants differ statistically significantly from the general population on psychological
variables. However, the large variation in effect sizes as well as the number of statistically
© Janine Lurie-Beck 2007
128
significant heterogeneity tests (Q homogeneity tests) suggest that differences BETWEEN
survivors and their descendants to the general population are not uniform. These findings
are suggestive of the notion that there is a large amount of heterogeneity WITHIN the
survivor and descendant populations.
The existing data pertaining to potential sources of this heterogeneity are examined
in Chapters Eight and Nine.
These chapters report the findings of meta-analytic
investigations into demographic sub-groups of the Holocaust survivor and descendant
populations.
© Janine Lurie-Beck 2007
129
Chapter Eight – Meta-Analyses of the Moderating Influence of Demographic Variables
among Holocaust Survivors
Chapter Four discussed numerous demographic variables which have been mentioned in
the literature as potential influences on Holocaust survivors’ post-war adjustment. This
chapter summarises the existing data pertaining to demographic differences within the
Holocaust survivor population with meta-analytic techniques being applied where possible.
Specifically, the demographic variables for which empirical assessment has been conducted
to date among the Holocaust survivor population, which are considered in this chapter are
the nature of Holocaust experiences, loss of family, gender, age during the Holocaust and
post-war settlement location. Meta-analysis techniques are used to provide additional and
in some cases new/first analyses of issues such as age during the Holocaust, post-war
settlement location, time lapse since the Holocaust and membership of survivor
organisations.
8.1. – Method
Direct and indirect assessment of demographic variables and their impact on Holocaust
survivors’ post-war adjustment are presented in this chapter.
Direct assessments of
demographic variables in the literature are summarised and meta-analysed where possible.
Indirect assessment, (where meta-analytic techniques are used to test variables not tested in
the individual studies included in the meta-analysis) is also possible for some variables
thereby filling some of the gaps in the literature.
Demographic variables are assessed indirectly via meta-analytic techniques such as
sub-set meta-analyses and meta-regressions. Sub-set meta-analyses provide insight into
demographic variables such as post-war settlement location via meta-analyses of studies in
sets determined by the country of study.
Meta-regressions are also used to explore
relationships between continuously-scaled variables and study effect sizes. Examples of
such continuous linear variables include the average age of study participants, and the
proportion of samples that meet particular criteria (such as the percentage of a sample that
is female).
8.2. – Nature of Holocaust Experiences.
Twenty studies were located that considered the influence of the nature or type of
Holocaust experiences on the presence or severity of post-war psychological symptoms.
The approaches used by researchers in the assessment of nature of Holocaust experiences
© Janine Lurie-Beck 2007
130
were so varied that it is impossible to conduct any valid meta-analyses. The dominant
operationalisation of this variable in the literature was categorical, but categories used are
not uniform across studies.
A number of studies use a camp versus non-camp
categorisation but the nature of these groups across studies differs widely. Some studies
use a severity rating which is subjectively determined and is also not consistent across
studies. The studies located which assessed the role of the nature of Holocaust experiences
are summarised in Table 8.1 with findings and trends outlined. While there are a few
exceptions to the rule, overall it appears that survivors with some form of camp experience
have suffered more detrimental effects than survivors with other experiences. It is clear
however that there are also differences depending on the kind of camp in which a survivor
was incarcerated (labour versus concentration versus death) and that conditions differed
between camps.
© Janine Lurie-Beck 2007
131
Table 8.1.
Summary of results from the literature based on the nature/type of Holocaust experiences endured by survivors
Study
Amir and LevWiesel (2001)
Operationalisation of Nature of
Experiences
Survivors who had been in hiding
during the war who could
remember their pre-war identity (n
= 23) and those who could not (n
= 23)
Ben-Zur and
Zimmerman
(2005)
30 concentration camp (CC)
survivors versus 30 survivors with
other Holocaust experience
(includes labour camps)
Brody (1999)
Camp (CC) versus non-camp
(NC) and also correlation with
length of imprisonment for camp
group
Clarke,
Colantonio,
Rhodes, Conn,
Heslegrave,
Links and van
Reekum (2006)
47 concentration camp survivors
(CC)
52 work camp, ghetto or hiding
survivors (WGH)
76 resistance fighters or other
Holocaust experience survivors
(OT)
26 camp survivors (C)
65 survivors who had been in
hiding (H)
Cohen, Dekel,
Solomon and
Lavie (2003)
Cordell (1980)
Favaro,
Rodella,
Colombo and
Santonastaso
(1999)
Correlations with length of
confinement in concentration
camp among 20 Holocaust
survivors
Italian political prisoners who had
been interned in a camp (n = 51)
(C)
Former partisans (n = 47) (P)
© Janine Lurie-Beck 2007
Findings
Trend
Depression and Anxiety scales of SCL-90
Depression – Could remember M = 0.72, Couldn’t remember
M = 1.08. This result narrowly missed achieving statistical
significance (t (44) = -1.91, p = 0.062).
Anxiety – Could remember M = 0.61, Couldn’t remember M
= 0.93
Negative Affect Scale
CC 3.26 versus Other 2.92
Statistically significant difference
Survivors who could not
remember their pre-war
identity more depressed and
more anxious
Geriatric Depression Scale, SCID-Depression, PTSD
Checklist for Civilians Intrusion, Avoidance and
Hyperarousal Subscales
GDS = CC 7.00, NC 6.67
SCID-D = CC 6.40, NC 6.13
Overall PTSD Severity = CC 32.60 NC 30.73
Intrusion = CC 1.73, NC 1.40
Avoidance = CC 1.47, NC 1.00
Hyperarousal = CC 1.47, NC 1.07
Positive correlation of r = 0.31 between PTSD symptom
severity and length of imprisonment for the camp group.
This correlation just missed out on statistical significance (p
< 0.06).
Some of the symptoms have higher incidence levels in the
non-camp group. Specifically these are flashbacks (20%
versus 7%), loss of interest in pleasurable activities (20%
versus 13%), emotional detachment (13% versus 7%),
foreshortened future (27% versus 13%), hypervigilance
(27% versus 20%) and exaggerated startle (30% versus
27%).
Geriatric Depression Scale, PTSD diagnosis
GDS = CC 18.2, WGH 18.9, OT 19.9
% with PTSD = CC 35%, WGH 27%, OT 13.2% (significant)
Camp survivors have higher
depression and PTSD
symptoms scores but some
individual symptoms are
more common among the
non-camp group with higher
levels of symptoms such as
hypervigilance and
exaggerated startle
potentially explained by
extended periods in hiding
Current PTSD symptom levels using the PTSD Inventory
Fear of Close Personal Relationships Questionnaire
PSTD symptom levels = C 7.65, H 5.95 Statistically
significant difference
FCPRQ – No statistically significant differences and no
descriptive data reported.
No statistically significant correlations between length of
confinement and paranoia and depression as measured by
Heimler Scale of Social Functioning. Correlation coefficients not quoted.
Incidence of major depressive disorder, depressed mood,
PTSD diagnosis, intrusion, avoidance and hyperarousal
Major depressive disorder = C 33%, P 4%
Depressed mood = C 55%, P 6%
PTSD Diagnosis = C 13%, P 2%
Intrusion = C 48%, P 28%
Avoidance = C 15%, P 6%
Hyperarousal = C 37%, P 28%
All differences are statistically significant using odds ratio
analysis
Among these symptom categories, the symptoms that
yielded the largest group differences were (in descending
order of significance) intense distress over reminders,
physiological reactivity over reminders and recurrent
nightmares (intrusion), avoidance of activities and situations,
avoidance of thoughts and feelings (avoidance) and
hypervigilance and exaggerated startle response
(hyperarousal)
Camp survivors suffer more
from PTSD symptoms than
survivors who were in hiding.
No trends can be noted about
relationship issues.
Concentration Camp
survivors more depressed
than non camp survivors and
labour camp survivors
Camp survivors have higher
PTSD diagnosis but no
significant difference in
depression severity
No trend can be noted due to
lack of data
Political camp prisoners more
depressed and more likely to
suffer from PTSD symptoms
than former partisans
132
Study
Hafner (1968)
Joffe, Brodaty,
Luscombe and
Ehrlich (2003)
Kuch and Cox
(1992)
Operationalisation of Nature of
Experiences
An examination of restitution
claim files.
Discrimination (experienced
effects of economic and social
laws imposed but emigrated
before enduring more extreme
experiences) = n = 95 (D)
Illegal residence (survivors in
hiding or using assumed identity)
= n = 70 (IR)
Ghetto = n = 54 (G)
Concentration camp = n = 158
(CC)
Mild experience group - generally
removed from high risk situations
such as living anonymously in the
countryside or with non-Jewish
families or living on Aryan papers
(n = 15)
Moderate group - usually in
ghettos or labour camps but not in
death camps, had some freedom
and were able to forage for food
(n = 39)
Severe group - in concentration or
death camps or in inhumane
conditions hidden for months and
often years, at constant risk of
being discovered or killed. (n =
46)
Subjectively determined by the
authors.
Review of 123 compensation files.
They delineated their sample into
concentration camp (n = 78) and
non-concentration camp (labour
camps or hidden) survivors (n =
45) and further divided their camp
sample into tattooed (Auschwitz)
(n = 20) and non-tattooed camp
survivors (n = 58).
Kuch, Rector
and SzacunShimizu (2005)
Review of 350 compensation files.
Ghetto versus tattooed camp
survivors
Leon, Butcher,
Kleinman,
Goldberg and
Almagor (1981)
Camp survivors (n = 27) versus
other (n = 15)
Letzter-Pouw
and Werner
(2005)
N = 96
Survivors of labour camps (LC),
concentration camps (CC) and
hiding (H) compared but no subsample sizes reported
Lev-Wiesel and
Amir (2000)
Child survivors - average 12 years
old in 1945.
Survivors who were in a
concentration camp (n = 35) (CC)
In hiding (n = 46) (H)
Adopted by a Christian family (n=
52) (A)
Cared for in a Christian
orphanage or monastery (n = 37)
(OM)
© Janine Lurie-Beck 2007
Findings
Trend
Incidence of Chronic depressive reactions and chronic
anxiety neurosis
Chronic depressive reactions = D 22%, IR 31%, G 19%, CC
30%
Chronic anxiety neurosis = D 19%, IR 41%, G 52%, CC 34%
Illegal residence and camp
highest on depression and
Illegal residence and ghetto
highest on anxiety
Severe Depression scale of the General Health
Questionnaire (SD-GHQ)
Withdrawn Depression scale of the Brief Psychiatric Rating
Scale (WD-BPRS)
Anxiety and Insomnia scale from the General Health
Questionnaire (AI-GHQ)
SD-GHQ = Mild 0.70, Mod 0.80, Sev 1.70
WD-BPRS = Mild 2.00, Mod 2.90, Sev 4.80
AI-GHQ = Mild 1.40, Mod 2.10, Sev 4.10
There are statistically significant differences for all three
scales.
Hiding survivors less anxious
and depressed than labour
camp or ghetto survivors who
are also less anxious and
depressed than
concentration/death camp
survivors.
PTSD diagnosis:
47% of the total sample met the criteria for diagnosis,
51% of the concentration camp sample
65% of the tattooed group.
The tattooed camp survivor group had a statistically
significantly higher number of PTSD symptoms as well (M =
9.40 versus M = 6.70). A list of each of the PTSD symptoms
is also provided and the tattooed camp survivor group
always had a higher incidence for each symptom than the
non-tattooed group.
Hamilton Depression and Anxiety Scales
Tattooed Auschwitz survivors did not score statistically
significantly differently from ghetto survivors on either scale.
No descriptive data was reported
Implication of differences in
PTSD rates depending on
severity of camp conditions
MMPI Depression and Paranoia Scales
Depression = CC 61.89, Other 64.13
Paranoia = CC 57.74, Other 55.87
NB – Means here were averaged across gender groups by
the current author to aid brevity
Impact of Events Scale – avoidance and intrusion subscales
Intrusion = LC 2.88, CC 3.82, H 3.05 (significant difference)
Avoidance = no significant difference and no means quoted
Depression, Anxiety, Phobic Anxiety and Paranoia scales of
SCL-90
PTSD Scale - Intrusion and Full or Partial PTSD (%)
Depression = CC 0.59, H 0.59, A 1.07, OM 0.78
Anxiety = CC 0.51, H 0.50, A 0.89, OM 0.76
Phobic Anxiety = CC 0.51, H 0.21, A 0.41, OM 0.27
Paranoia = CC 0.34, H 0.53, A 0.72, OM 0.77
Intrusion = CC 3.86, H 2.88, A 3.75, OM 4.38
Full or Partial PTSD Diagnosis = CC 49%, H 48%, A 50%,
OM 19%
All scale scores differ statistically significantly across the
groups
No statistically significant
difference between ghetto
and camp survivors but no
trend can be noted due to
lack of data.
Survivors with non-camp
experiences such as hiding
have higher depression and
lower paranoia than camp
survivors
Concentration camp
experience associated with
highest intrusion levels
Survivors adopted or cared
for in a Christian institution
have higher levels of
depression, anxiety and
paranoia than survivors who
hid or were in camps.
Survivors in institutions have
notably lower level of PTSD
diagnosis but have highest
intrusion rate.
133
Study
Nathan,
Eitinger and
Winnik (1964)
Operationalisation of Nature of
Experiences
Sorted through Jerusalem’s
Talbieh Psychiatric Hospital’s
patient records for the period
1949 to 1959 and found the files
of 157 concentration camp
survivors and 120 survivors who
had spent most of the war in exile
in the Soviet Union.
Findings
Trend
Incidence of major depressive disorder, dysphoria,
depressive signs, anxiety attacks, free floating anxiety,
paranoid manifestations and paranoid diagnosis
MDD = C 13%, E 10%
Dysphoria = C 45%, E 21% Statistically significant difference
Depressive signs = C 49%, E 37% Statistically significant
difference
Anxiety attacks = C 3%, E 1%
Free floating anxiety = C 15%, E 5% Statistically significant
difference
Paranoid manifestations = C 41%, E 43%
Paranoid Diagnosis = C 9%, E 8%
The reader should bear in mind however that the camp
group appear to have a slightly older age profile than the
exile group, meaning age could be a possible confound in
this study.
Depression and anxiety diagnoses
Depression diagnosis = C 55%, NC 24%
Anxiety diagnosis = C 55%, NC 27%
Both differences are statistically significant when analysed
via odds ratio analysis
Camp survivors have higher
incidence of depression,
anxiety and paranoia than
people who spent time in
exile in the USSR
Child survivors who had been in
camps (n = 43) versus those who
had been in hiding (n = 44)
Depression Diagnosis
Camp group 48%
Hiding group 31%
Not statistically significant
Camp group more likely to be
diagnosed with depression
than hiding group.
Concentration camp group (n =
47) (C) versus hiding group (n =
53) (H)
Beck Depression Inventory and Spielberger Trait-Anxiety
Scale.
BDI = CC M = 10.02, H = 11.45
STAS = CC M = 41.64, H = 43.66
Impact of Events Scale
Based on a study whose main interest was differences
before and after open heart surgery among a group of
Holocaust survivors. The before surgery scores are
reported here. These results are of course affected by the
heart problems obviously being experienced by the survivor
sample.
Total PTSD score = C 17.00, GH 16.40, E 17.40
Intrusion = C 9.60, GH 8.60, E 12.90 Statistically significant
difference between GH and E
Avoidance = C 6.80, GH 8.40, E 4.50 Statistically significant
difference between GH and E
Impact of Events Scale
Correlations between avoidance and intrusion symptoms
and length of confinement in ghettos and/or camps. Minimal
relationship between these variables was found with the
highest correlation being 0.19 between avoidance and
number of months confined in camps.
Intrusion = A 25.72, C/G 27.39
Avoidance = A 16.00, C/G 14.54
None of the 17 symptoms listed on the Clinician
Administered PTSD Scale statistically significantly
differentiated between their sample of survivors who had
been in hiding and camp survivors. No descriptive data
were reported.
Hiding group more depressed
and anxious than camp
group.
Robinson,
Rapaport,
Durst,
Rapaport,
Rosca, Metzer
and Zilberman
(1990)
Robinson,
Rapaport-BarSever and
Rapaport
(1994)
Rozen (1983)
49 death camp survivors (C)
versus
37 non-camp survivors (NC)
Schreiber,
Soskolne,
Kozohovitch
and Deviri
(2004)
Camp survivors (n = 25) (C)
Ghetto/hiding survivors (n = 25)
(GH)
Evacuated before experiencing
the whole gamut of possible
traumas (n = 13) (E)
Silow (1993)
Correlations with length of
confinement in ghettos and/or
camps (whole sample n = 38)
Auschwitz survivors (n = 25) (A)
Survivors of other camps/ghettoes
(n = 13) (C/G)
Yehuda,
Schmeidler,
Siever, BinderBrynes and
Elkin (1997)
Camp (n = 70) versus hiding (n =
30)
Camp survivors more likely to
be diagnosed with depression
or anxiety than non-camp
survivors
Seemingly incongruous result
that survivors who only
experienced early
persecution phase are more
affected by PTSD symptoms
but this study is affected by
sample with heart problems
awaiting surgery
Varied results for each PTSD
symptom cluster when
comparing Auschwitz to other
survivors
No trend can be noted due to
lack of data
8.3. – Country of Origin
Only one study was located that addressed the potential influence of a survivor’s country of
origin on their symptom levels. Letzter-Pouw and Werner (2005) report that survivors who
were born in Eastern Europe reported significantly higher levels of intrusion (M = 3.54, n =
78) than survivors born in Western Europe (M = 2.43, n = 18, F (1,85) = 15.9, p < 0.01).
© Janine Lurie-Beck 2007
134
Note that no standard deviations were reported. Avoidance symptoms were also recorded
but no details were reported as no significant difference was found.
8.4. – Loss of Family Members
Four studies were located that considered the effect of the loss of family members during
the Holocaust. Each of these studies operationalised their analysis in a different way and so
meta-analytic methods could not be applied with this demographic variable.
Hafner (1968) cross-referenced survivors who were diagnosed with depression,
anxiety or paranoia with those who had lost members of their immediate family during the
war. The incidence of recurrent depression was slightly higher among those who had lost
family members than among those who had not. However, contrary to what would be
expected, chronic depression was markedly higher among those who had not lost family
members. This data was provided only in a chart without exact numbers quoted however it
is possible to state that approximately 18% of survivors without loss (n = 107) were
diagnosed with recurrent depression compared to approximately 24% who did (n = 106)
and approximately 26% of survivors with loss were diagnosed with chronic depression
compared to approximately 38% of those who did not lose family members. The incidence
of free floating anxiety was higher among those who had not lost family members than
among those who had. Approximately 40% of survivors without loss (n = 107) were
diagnosed with free floating anxiety compared to approximately 28% who did (n = 106).
Though it came close, this difference did not reach statistical significance when assessed
using odds ratio analysis. Hafner (1968) theorised that the higher level of anxiety and
depression among survivors who did not lose family members may be due to interaction
with family members also traumatised which may be stronger than the impact of losing
family members. Hafner (1968) also compared paranoid reactions among survivors who
had not lost any family members to those who had lost both their parents and other
members of their immediate family. Among the survivors with no loss, only one had a
paranoid reaction. Among those who had lost family members, 4 or 4% had evident
paranoid reactions. The odds ratio, as calculated by the current author, is 4.16 (in favour of
survivors who had lost family members) but does not reach statistical significance because
of the small proportions involved.
In his thesis, Silow (1993) divided his sample of 38 Jewish Holocaust survivors into
three groups based on the number of family members they had lost during the war. The
three groups were: 1) Sole survivors, 2) Survivors with 2-3 surviving relatives (including
© Janine Lurie-Beck 2007
135
self) and 3) Survivors with 4-18 surviving relatives.
Two one-way ANOVAs were
conducted to assess the differences between these three groups on intrusion and avoidance.
Neither statistical test reached significance. Unfortunately no means or standard deviations
were quoted in order to note even a small trend in any direction.
In another unpublished thesis, Brody (1999) ran a correlation analysis between the
number of family members who were killed during the Holocaust and severity of PTSD
symptoms.
She included her control group within this analysis too.
These were
participants who had not directly experienced the Holocaust but who nonetheless did lose
relatives in the Holocaust. She obtained a statistically significant positive correlation of r =
0.35 (p < 0.03) between severity of PTSD symptoms and the number of family members
killed during the Holocaust for the total sample of 40 participants (including 10 control
participants).
Finally, in another unpublished thesis, Cordell (1980) asserts she found no
statistically significant difference in depression or paranoia between survivors who were
alone and survivors who had family with them while in a concentration camp.
Unfortunately, she quotes no means so that not even a slight trend can be discerned from
her results.
Results relating to the impact of loss of family members are certainly difficult to
interpret. Unreported descriptive data and contaminated samples make it difficult to truly
gauge the impact of this variable in three of the four studies. The findings from Hafner’s
(1968) study seem counterintuitive. This variable has not yet been analysed adequately to
make any firm statements about its true influence.
8.5. – Gender
The influence of gender on a survivor’s post-war adjustment is analysed a number of ways.
Firstly, results that directly compared male and female survivors are meta-analytically
combined in effect size and odds ratio meta-analyses. In addition, an indirect method is
used in which studies comparing survivors to control groups are analysed according to the
female percentage of their sample. Studies included in this meta-regression are delineated
in Appendix A. A comparison of male versus control and female versus control is also
provided (see Appendix D for details of studies included).
8.5.1. – Meta-analytic Results
An examination of the meta-analytic findings presented in Tables 8.2, 8.3 and 8.4 indicate
that overall it is female Holocaust survivors who appear more negatively affected by
© Janine Lurie-Beck 2007
136
Holocaust trauma than male survivors. In addition, there is tentative support for the idea
that not only are females more affected than males but that female survivors differ more
widely from female controls than male survivors do from male controls.
Given that
females in general have been found to suffer from higher rates of psychological symptoms
such as those tested here, (for example Oltmanns & Emery, 1995) these results hint that the
gender differences in the survivor population are over and above the gender differences one
might expect in the general population.
Table 8.2.
Meta-analysis of survey study results based on survivor gender
Variable
Depression
Anxiety
Paranoia
Intrusion
Avoidance
Effect size for male versus female results
Number of
results where
females
scored higher
than males
3/5
1/1
2/2
1/1
0/1
g
95%
confidence
limits for g
Total
N
- 0.39 *
- 0.65 *
- 0.53
- 0.16
0.08
- 0.63 to - 0.15
- 1.05 to - 0.24
- 1.16 to 0.11
- 0.80 to 0.48
- 0.56 to 0.72
324
100
42
38
38
Effect size for male versus control
male results
95%
Survivor
g
confidence
N
limits for g
0.18
0.65
- 0.15 to 0.51
No results located
- 0.05 to 1.34
No results located
No results located
76
17
Effect size for Females versus control
female results
95%
Survivor
g
confidence
N
limits for g
0.32
0.30
0.73 *
- 0.03 to 0.67
- 0.10 to 0.70
0.15 to 1.32
No results located
No results located
57
48
25
Notes. A negative effect size indicates that females scored higher than males. Results for paranoia versus controls
based on 1 study with 2 male and female samples. Results for anxiety females versus controls based on 1 study with 2
measures of anxiety (averaged effect size). Results for depression versus controls based on 3 results from 2 studies.
* p < 0.05
Table 8.3.
Meta-regression of survivor versus control results with the female percentage of the survivor sample
Variable
Number of results for which
gender breakdown was
provided
24/30
8/13
8/12
3/9
3/9
3/3
3/3
Depression
Anxiety
Paranoia
Assumption that the World is Benevolent
Assumption that the World is Meaningful
Positive Attachment Dimensions
Negative Attachment Dimensions
rmeta
2
rmeta
Survivor group
sample size
0.23 ***
- 0.30 ***
0.12
- 0.86 ***
0.77 ***
- 0.98 ***
0.97 ***
0.05
0.09
0.01
0.75
0.59
0.95
0.95
1,480
378
160
178
178
130
130
Note. Study results included in meta-regressions are weighted in the analyses by their associated survivor sample size
so that results based on larger samples are weighted more heavily to reflect their increased precision in estimating the
true population effect.
*** p < 0.001
Table 8.4.
Summary of meta-analyses of incidence/diagnosis studies comparing male and female survivors
Variable
Depression
Anxiety
Paranoia
PTSD #
Number of results where
females had higher
incidence than males
2/2
2/2
1/1
0/1
Ω
95% confidence
3.17 *
2.52 *
2.90 *
1.61
2.65 to 3.81
2.11 to 3.01
2.21 to 3.82
0.56 to 4.64
limits for
Ω
Average
incidence
among females
49%
68%
18%
19%
Average incidence
among males
Survivor group
sample size
23%
45%
7%
27%
2,303
2,302
2,159
91
Note. The odds ratio reported for PTSD is in reference to males rather than females. In other words the odds of males
having PTSD in this study sample are 1.61 times higher than the odds for the females in the sample. All other odds ratios
reported here indicate a higher incidence for females.
Ω meta-analytic odds ratio
* p < 0.05
© Janine Lurie-Beck 2007
137
Table 8.5.
Meta-regression of incidence rates among survivor with the female percentage of the survivor sample
Variable
Number of results for
which gender
breakdown was provided
Depression
Anxiety
Paranoia
5/5
4/4
2/2
rmeta between
incidence and
female % of
sample
0.25 ***
0.67 ***
1.00 ***
2
rmeta
Survivor group
sample size
0.06
0.45
1.00
2,462
2,418
2,209
Note. Study results included in meta-regressions are weighted in the analyses by their associated survivor sample size
so that results based on larger samples are weighted more heavily to reflect their increased precision in estimating the
true population effect.
*** p < 0.001
8.5.2. – Studies Excluded from Meta-analyses
Children of Holocaust survivors provided ratings of their male and female survivor parents
in three studies. Because they are ratings by third parties and not self-ratings they are not
included in the meta-analyses.
Podietz et al. (1984) asked a group of children of survivors to rate their survivor
parents in terms of how fearful they perceived them to be. The researchers did not cite
group means and standard deviations but do report that children of survivors rated their
mothers and their fathers as statistically significantly more fearful than control mothers and
fathers were rated (t (216) = 3.63, p < 0.001 and t (216) = 2.58, p < 0.05 respectively). The
difference between the groups was more marked for ratings of mothers, however because of
lack of information the difference between how survivor mothers and fathers were rated
could not be directly assessed. The raw data would have been required to conduct a
repeated measures t-test to discover if children of survivors rated their mother statistically
significantly higher than their fathers.
On the anxious/nervous scale of the Adjective Checklist a group of 73 children of
survivors rated their survivor mothers statistically significantly higher than their survivor
fathers in a study by Leventhal and Ontell (1989). Survivor mothers were given a mean
rating of 3.79 (SD = 1.11) compared to 3.39 (SD = 1.30) given to fathers. This difference
was statistically significant at the 0.05 level (t (144) = -1.99, p < 0.05). Major (1996) asked
some children of survivors and a control group to rate their parent’s on a variable they
termed suspiciousness. Both groups rated their mothers and fathers very similarly (in fact
the mean ratings assigned ranged only between 2.4 and 2.5. Obviously such differences did
not reach statistical significance. The control group (n = 37) rated both their mother and
their father as 2.5 on this questionnaire while survivors’ children (n = 19) rated their fathers
as 2.4 and mothers at 2.5.
© Janine Lurie-Beck 2007
138
In the third study to present the perceptions of Holocaust survivors held by their
children, Woolrich (2005) reported the percentage of survivors’ children who considered
their survivor parents to be anxious or depressed. Of the 59 participants with a survivor
mother, 27% considered her depressed and 29% considered her anxious. Of the 69 with a
survivor father, 26% rated their father as depressed and 21% considered him to be anxious.
While the differences in perceived depression were slight, survivor mothers were more
likely to be considered anxious than survivor fathers.
A number of studies searched for gender differences in their combined data set of
Holocaust survivors and control participants. There is of course no way of telling whether
the findings obtained would have been stronger or weaker had the survivor sample been
partitioned out but it is still interesting to note their results. In a study by Prager and
Solomon (1995) a regression analysis was undertaken with a number of variables including
gender. Gender was found to be a statistically significant predictor for the meaningfulness
of the world subscale in the combined Holocaust survivor and control participant sample.
Specifically, females scored lower than males on this subscale. No descriptive statistics
were provided by gender to allow for significance testing of gender within the Holocaust
survivor sample. Another study which considered gender differences on world assumptions
was that by Breslau (2002). Again, it is unfortunate that the analysis was based on control
and survivor groups pooled together, however her results are similar to those obtained by
Prager and Solomon (1995). Specifically she found no specific gender differences for the
Self-Worth or Benevolence of the World subscales but found that females scored
statistically significantly lower than males on the Meaningfulness of the World subscale.
The mean score for females was 36.93 compared to 39.75 for males (F (1,251) = 6.80, p <
0.01). Finally, via multiple regression analysis in which gender was a predictor, Brody
(1999) asserted that it was females who suffered statistically significantly more from both
intrusion and avoidance than males in her combined survivor and control sample. While
the results from all three studies were tainted by the use of the combined control group and
survivor sample, it is worthy of note that overall their findings do support the prevailing
trend of females being worse off than males.
Two additional studies to consider gender differences could not be included in
meta-analyses because of insufficient data.
Yehuda et al. (1997) found statistically
significant differences between two symptoms within the avoidance cluster. The nonstatistically significant gender differences on all other symptoms were not reported so no
© Janine Lurie-Beck 2007
139
general trend can be ascertained.
They also did not report on overall avoidance or
intrusion results and so this study could not be included in the meta-analysis of PTSD
symptoms. In terms of the two statistically significant differences, female survivors (n =
67) scored statistically significantly higher than male survivors (n = 33) on psychogenic
amnesia (M = 2.55, SD = 2.63 versus M = 0.85, SD = 1.64, t (98) = 3.37, p < 0.001) and
foreshortened future (M = 1.73, SD = 2.30 versus M = 0.82, SD = 0.88, t (98) = 1.76, p <
0.05). Gender was a non-statistically significant predictor of post-traumatic growth in LevWiesel and Amir’s (2003) study. Unfortunately, no means were reported by gender and so
no trends can be described.
8.6. – Age during the Holocaust
The influence of a survivor’s age during the Holocaust on their post-war psychological
health was assessed by thirteen studies and also indirectly by the current author via metaregression. Studies directly assessing age, operationalised it either as a continuous variable
or as a categorical variable with various age cut-offs. The results of these studies are
summarised in Tables 8.6 and 8.7. Across the studies there are a few findings that are
ambiguous, however, overall, there is a trend towards negative effects increasing rather
than decreasing with age.
© Janine Lurie-Beck 2007
140
Table 8.6.
Studies assessing impact of survivor age via correlation/regression analysis
Study
Brody (1999)
Cohen,
Dekel,
Solomon and
Lavie (2003)
Cordell
(1980)
Lev-Wiesel
and Amir
(2003)
Prager and
Solomon
(1995)
Schreiber et
al. (2004)
Yehuda at al
(1997)
Findings
The information provided on the impact of age on PTSD
symptoms was only calculated for the entire sample (including
both survivors and controls). A correlation of r = 0.05 between
age and PTSD severity was obtained. It is reported (without
associated descriptive statistics) that older subjects suffered
more from intrusion and avoidance (as opposed to hyperarousal)
than younger subjects. Again this was not delineated by group.
It cannot be determined whether a stronger correlation would
have been derived if just based on the survivor sample.
There was a small (not-significant) negative correlation of -0.23
between age in 1939 and fear of intimacy for 43 treated
survivors, with a null-result of 0.04 for 48 non-treated survivors.
The correlation between age and current PTSD symptoms was r
= 0.20 for the treated group r = 0.32 (p < 0 .05) for the nontreated group.
No statistically significant correlation between either the
depression or paranoia subscales of the Heimler Scale of Social
Functioning and survivor age during the Holocaust in a group of
20 Holocaust survivors.
Survivors’ age was a statistically significant predictor of overall
post-traumatic growth. The beta weight quoted was a positive
number (β = 0.24) suggesting that that the older the survivor the
higher their post-traumatic growth score.
Age was not statistically significantly correlated with any of the
three subscales of the World Assumptions Scale. No coefficients were cited.
The main focus of study was on PTSD levels before and after
heart surgery. A statistically significant negative correlation was
found between pre-surgery avoidance scores and age (r = -0.30,
p < 0.05) among the Holocaust survivor group.
Derived a positive correlation between overall PTSD symptom
levels and age among their sample of 100 survivors (r = 0.61, p <
0.001). When the symptoms were examined individually it was
found that this effect was not uniform. Specifically, negative
correlations with age were found for the PTSD symptoms of
hypervigilance (p = 0.001), psychogenic amnesia (p = 0.008) and
emotional detachment (p = 0.046). Distressing intrusive thoughts
was the only individual symptom to have a statistically significant
positive correlation (p = 0.013). Correlation co-efficients were
not reported for these individual symptoms. Clearly the majority
of the remaining symptoms all had relationships in the positive
direction to create the overall statistically significant positive
effect.
© Janine Lurie-Beck 2007
Trend
Mixed/unreliable findings
PTSD symptoms increase with age
No effect.
Post-traumatic growth higher for older survivors
No effect
PTSD avoidance decreases with age but biased data
Overall PTSD symptoms increase with age but a few specific
symptoms decrease with age
141
Table 8.7.
Studies assessing the impact of survivor age categorically
Study
Bower (1994)
Nature of Age Categories
Comparison of survivors aged 16 or
younger (n = 100) during the Holocaust
to survivors aged 20 years or over (n =
100)
Findings
Aged 16 or younger group = 66% depression, 55% anxiety, 29%
‘contact abnormalities’ which includes paranoid ideation.
Aged 20 years or over = 76% depression, 52% anxiety, 10%
‘contact abnormalities’ which includes paranoid ideation.
The difference between the groups on contact abnormalities is
statistically significant (odds ratio = 3.68, 95% CI = 1.68 to 8.05)
Brom, Durst
and Aghassy
(2002)
Clients of AMCHA (an organisation
aimed at providing psychological help to
Holocaust survivors).
AMCHA adult group (n = 60) = average
age of 75 at the time of the study and
were therefore aged 21 on average at
the end of the war
AMCHA child group (n = 28) = average
age of 64 and therefore were aged 10
on average at the end of the war.
Differing categories for each analysis.
Child group = Total IES Score (M = 33.60, SD = 14.30), Avoidance
(M = 13.90, SD = 10.00), Intrusion (M = 19.10, SD = 10.80), lower
on benevolence and meaningfulness scales of the World
Assumptions scale (rescaled from the 8 sub-scale version)
Adult group = Total IES Score (M = 31.00, SD = 13.20), Avoidance
(M = 10.30, SD = 7.30), Intrusion (M = 20.70, SD = 10.20), higher on
benevolence and meaningfulness scales of the World Assumptions
scale (rescaled from the 8 sub-scale version)
No statistically significant differences.
Hafner (1968)
Keilson and
Sarphatie
(1992)
Age when separated from mother during
the Holocaust.
0 to 18 months (n = 30)
13 months to 4 years (n = 41)
4 to 6 years (n = 24)
6 to 11 years (n = 42)
11 to 14 years (n = 26)
14 to 18 years (n = 41)
Review of compensation files of 63
survivors aged under 30 at the start of
their incarceration in a camp versus 81
survivors who were over 30 at the time
Matussek
(1975)
Robinson,
Rapaport,
Durst,
Rappaport,
Rosca,
Metzer, and
Zilberman
(1990) and
Robinson,
Rapaport-BarSever and
Rapaport
(1994)
Two studies deriving sample from the
same source (Yad Vashem records)
1994 study group restricted to survivors
aged less than 13 (n = 103) when
persecution began
1990 study group contained adult
survivors (n = 86)
Trend
Depression
increasing with
age
No effect for
anxiety
Paranoia
decreasing with
age (statistically
significant result)
PTSD score
decreasing with
age
World
Assumptions
more positive
with increasing
age
Aged 13 at start of persecution (n = 37) = 9% chronic depression
diagnosis, 60% anxiety neurosis and other neurotic reactions
diagnosis, and 6% paranoid reactions
Aged 14 to 21 (n = 104) = 24%, 59%, and 1%
Aged 22 to 30 (n = 98) = 26%, 59% and 1%
Aged 31 to 50 (n = 92) = 40%, 49% and 1%
Aged 51 and over (n = 4) = 50%, 50% and 0%
Approximations from a graph (specific data not cited)
Aged 0 to 8 years (n = 12) = 15% symptom of depression, 31%
symptom of free floating anxiety
Aged 9 to 13 years (n = 26) = 26% and 39%
Aged 14 to 16 years (n = 33) = 31% and 19%
Aged 17 to 21 years (n = 68) = 35% and 39%
Aged 22 to 30 years (n = 93) = 40% and 25%
Aged 31 to 50 years (n = 88) = 39% and 38%
Aged 51 and over (n = 4) = 50% and 50%
0 to 18 months = 7% chronic-reactive depression diagnosis, 13%
anxiety-neurotic development diagnosis
13 months to 4 years = 12% and 27%
4 to 6 years = 8% and 25%
6 to 11 years = 19% and 12%
11 to 14 years = 12% and 39%
14 to 18 years = 32% and 20%.
Depression
increasing with
age
Mixed results for
anxiety
Paranoia
incidence too
low to note a
pattern
Aged under 30 = 32% depressive mood, 33% permanent anxiety
state
Aged over 30 = 30% and 38%
Neither of these differences were statistically significant using odds
ratio analysis.
Younger group = 42% depression, 43% anxiety
Older group = 42% depression, 44% anxiety
No statistically significant differences.
Little difference
for depression
Anxiety
increasing with
age
No effect
Mixed findings
for anxiety and
depression
Meta-regressions are conducted using the effect sizes calculated for studies
comparing Holocaust survivors to control groups (as presented and summarised in Chapter
Seven) and the survivor samples’ average age in 1945. The mean age of each study group
© Janine Lurie-Beck 2007
142
in 1945 is calculated by subtracting the mean age from the study year to determine the
mean birth year and then subtracting this from 1945. Not every study provided the average
age of their sample and so the complete set of studies/results cannot be included in the
meta-regressions. The proportion of results that can be included in these analyses is
outlined in Table 8.8 below. The details of all studies included in the meta-regression and
the studies that could not be included can be seen in Appendix A.
Table 8.8.
Summary of meta-regression findings for average age of survivors in 1945
Variable
Depression
Anxiety
Paranoia
Assumption that World is Benevolent
Assumption that World is Meaningful
Positive Attachment Dimensions
Negative Attachment Dimensions
Number of results with age
data provided which could be
included in the metaregression
16/30
8/13
2/12
3/9
3/9
3/3
3/3
rmeta
2
rmeta
Survivor group
sample size
- 0.54 ***
0.26 ***
-1.00 ***
- 0.89 ***
0.58 ***
0.65 ***
- 0.25 **
0.29
0.07
1.00
0.80
0.33
0.42
0.06
1,313
376
53
178
178
130
130
Note. Study results included in meta-regressions are weighted in the analyses by their associated survivor sample size
so that results based on larger samples are weighted more heavily to reflect their increased precision in estimating the
true population effect.
** p < 0.01, *** p < 0.001
The results from the meta-regressions provide some contrasting findings. The data
pertaining to depression, world meaningfulness and negative and positive attachment
dimensions suggest that negative effects decrease with age. The results for anxiety and
world benevolence are more consistent with the overall pattern from the studies that
directly assessed age, namely that older survivors suffered more ill-effects. Obviously the
result for paranoia is remarkable but its perfect magnitude is likely a statistical aberration.
The findings of the one study found that reports on children of survivors’ perception
of their parents, suggested ill-effects increasing with survivor age. Sigal and Weinfeld
(2001) had Canadian children of Holocaust survivors rate their parents on a number of
psychopathological measures including depression. These children were separated into
four groups based on their parents’ age at the cessation of World War II. These groups
consisted of children of survivors aged 2-9 years (n = 47), 10-13 years (n = 54), 14-18
years (n = 114) or 19-24 years (n = 137) at the end of the war. In general, older survivors
were perceived to have higher depression and paranoia levels than younger survivors by
their children. Survivors aged 2-9 years during the war were rated as the least depressed
(M = 2.35, SD = 1.87), followed by those aged 10-13 years (M = 2.63, SD = 2.29) with the
14-18 and 19-24 year olds being almost equivalent (M = 3.64, SD = 2.53 and M = 3.63, SD
© Janine Lurie-Beck 2007
143
= 2.66, respectively). The two older age groups were rated statistically significantly higher
than the two younger age groups by their children. Survivors 2-9 years were rated as the
least paranoid (M = 1.16, SD = 1.54), followed by those aged 10-13 years (M = 1.43, SD =
1.61), 14-18 years (M = 1.96, SD = 1.86), and finally 19-24 year olds (M = 2.27, SD =
1.78). Children of survivors who were 2-9 years old during the war rated their parents’
paranoia as statistically significantly lower than those whose parents were 14-24 years old.
In addition, children whose parents were 10-13 years also rated their parents’ paranoia as
statistically significantly lower than the 19-24 years group.
The findings in relation to survivors’ age during the Holocaust are far from clear. It
is not discernible whether conflicting results are due to differing methodology or analysis
approaches or some other reason. Further analysis of the impact of age is certainly
warranted, preferably with age measured on a continuous scale rather than as a categorical
variable.
8.7. – Time Lapse since the Holocaust
Two meta-regression sets are calculated to determine the relationship between the time that
has elapsed since the Holocaust and survivors scores on and incidence levels of depression,
anxiety and paranoia (as compared to control groups). Effect sizes and incidence rates are
correlated with the time that had elapsed between 1945 and when each study was
conducted. Studies included in this meta-regression can be seen in Appendix A. No
studies were excluded.
Table 8.9 reports the results of the meta-regressions based on the survey studies.
The results for depression and anxiety are statistically significant quadratic correlations
which means they reflect curvilinear rather than linear relationships. These curvilinear
relationships can be seen more clearly in Figure 8.1. As can be seen, these analyses suggest
that depression and anxiety symptoms have shown an upward trend in more recent years
after having subsided somewhat. The linear meta-regression correlation for paranoia is also
statistically significant but differences between linear and curvilinear analyses are not
significant. It is not clear whether there is truly a linear relationship between time lapse and
paranoia or whether there is a curvilinear relationship that has been left uncovered because
of smaller sample sizes.
© Janine Lurie-Beck 2007
144
Table 8.9.
Summary of meta-regression findings for time lapse since the Holocaust among survey studies
Variable
Number of results included
in meta-regression
30
13
12
Depression
Anxiety
Paranoia
rmeta (linear)
2
(linear)
rmeta
rmeta (quadratic)
2
(quadratic)
rmeta
0.39 ***
0.50 ***
0.37 ***
0.16
0.25
0.14
0.57 ***
0.51 **
0.37
0.33
0.26
0.14
Survivor group sample
size
2,000
907
360
Notes. Study results included in meta-regressions are weighted in the analyses by their associated survivor sample size
so that results based on larger samples are weighted more heavily to reflect their increased precision in estimating the
true population effect.
*** p < 0.001
1.50
1.50
Effect Size for Anxiety
Effect Size for Depression
2.00
1.00
0.50
1.00
0.50
0.00
0.00
-0.50
-0.50
1980.00
1985.00
1990.00
1995.00
2000.00
2005.00
1975.00
1980.00
1985.00
1990.00
1995.00 2000.00
2005.00
Year of Study
Year of Study
Figure 8.1. Scatterplots of effect sizes comparing survivor groups to control groups on depression and anxiety surveys
and the year studies were conducted
Note. Each study is weighted in the analysis by its survivor sample size. This weighting is denoted by the size of the
data points in the Scatterplot.
Table 8.10 reports the meta-regressions for incidences of depression, anxiety and
paranoia among survivors reported and time lapse since the war. There appears to be a
decrease in depression and paranoia diagnosis but an increase in anxiety diagnosis with the
passage of time.
However it is unclear whether these statistically significant linear
relationships actually hide curvilinear relationships that would be more apparent with larger
study sets.
Table 8.10.
Summary of meta-regression findings for time lapse since the Holocaust among incidence studies
Variable
Depression
Anxiety
Paranoia
Number of results included in
meta-regression
6
6
3
rmeta
2
rmeta
-0.71 ***
0.66 ***
-0.85 ***
0.50
0.44
0.72
Survivor group
sample size
2,606
3,444
2,353
Notes. Study results included in meta-regressions are weighted in the analyses by their associated survivor sample size
so that results based on larger samples are weighted more heavily to reflect their increased precision in estimating the
true population effect. It was not possible to validly and reliably test for quadratic/curvilinear relationships due to the small
study sets involved.
*** p < 0.001
© Janine Lurie-Beck 2007
145
Two studies conducted by Robinson with various colleagues also provide data
pertinent to this issue by providing incidence of depression and anxiety as at the time of
study and also retrospectively immediately after the war (Robinson et al., 1994; Robinson
et al., 1990).
In the earlier study 55% of the 86 survivors suffered from depression
immediately after liberation compared to 42% in 1989 when the data was collected. They
further note that 38% had suffered with depression for the entire time since liberation, 17%
endured it only after the war but not currently but 4% said they only suffer from depression
currently and didn’t suffer from it after the war. In terms of anxiety, 47% said they had
suffered from anxiety in the post-war period and 44% said they suffer from anxiety
currently. Over a third (35%) said they had suffered from anxiety continuously to the
present day since the Holocaust while one in ten (11%) said they did suffer from anxiety in
the immediate post-war period but no longer suffer from it. Interestingly though just under
one in ten (8%) said that they did not suffer from anxiety in immediate aftermath of the war
but were experiencing symptoms at the time of the study. On a positive note, the authors
note that close to half of the sample surveyed (46%) said that they have not experience
anxiety at all (either immediately after the war or since (Robinson et al., 1990).
In the later study conducted with child survivors (aged less than 13 when
persecution began) the incidence of depression was higher when the data was collected than
after the war (31% versus 42% of 103 survivors). A higher proportion of those surveyed
indicated that they suffer from chronic anxiety today than said that they experienced such
anxiety after the war (43% versus 37% of 103). This would suggest that there is a group of
survivors within that sample who did not experience anxiety in the post-war period but later
developed it.
8.8. – Post-war Settlement Location
A survivors’ post-war settlement location had/has the potential to affect and be affected by
a survivors’ post-war adjustment. The way survivors were received by society differed by
country but also the reasons a survivor had for choosing to settle in a particular country also
differed. Survivors who were the most affected by their Holocaust experiences may have
been the ones who chose to settle the furthest away from Europe. Differential results
dependent on post-war settlement location are presented in this sub-section. Only two
studies that addressed this issue directly could be found. Sub-set meta-analyses of studies
grouped together by study location provide an indirect measure of this variable. The results
© Janine Lurie-Beck 2007
146
of meta-analytic sub-set analysis of survey and incidence/diagnosis studies are presented in
Tables 8.11 and 8.12. Details of the studies included can be found in Appendix A.
Table 8.11.
Summary of sub-set meta-analyses of survey studies by post-war settlement location among survivors
Anxiety
Depression
Paranoia
Assume world
is benevolent
Assume world
is meaningful
Number of
results based
on data
collected in
Israel
7
14
4
6
6
Israel
Survivor
group
sample
size
America, Australia or Canada
Number of results
Survivor
g
based on data
group
collected in America,
sample
Australia or Canada
size
g
95%
confidence
limits for g
95%
confidence
limits for g
329
1,197
249
249
0.62 *
0.29 *
0.33 *
-0.28 *
0.39 to 0.85
0.21 to 0.37
0.15 to 0.51
- 0.41 to - 0.15
6
16
8
1
501
803
411
67
0.53 *
0.37 *
0.52 *
-0.30
0.39 to 0.68
0.27 to 0.47
0.38 to 0.66
- 0.63 to 0.04
249
-0.09
- 0.22 to 0.03
1
67
-0.25
- 0.58 to 0.09
* p < 0.05
Table 8.12.
Summary of sub-set meta-analyses of incidence/diagnosis studies by post-war settlement location among survivors
Variable
Anxiety
Depression
Number
of results
based on
data
collected
in Israel
Ω
3
4
1.41 *
1.24 *
Israel
95%
confidence
limits for
Average
Incidence
Ω
1.24 to 1.62
1.07 to 1.43
55%
34%
Survivor
Group
Sample
Size
2,286
2,327
Number
of results
based on
data
collected
in
Canada
1
1
Ω
Canada
95%
confidence
limits for
Average
Canada
Incidence
Survivor
Group
Sample
Size
30%
18%
135
135
Ω
1.35
1.40
0.78 to 2.32
0.72 to 2.72
Ω meta-analytic odds ratio
* p < 0.05
The meta-analytic results based on incidence studies suggest higher incidence of
both anxiety and depression among survivors who settled in Israel compared to survivors
who settled in Canada (see Table 8.12). The results from the sub-set meta-analyses of
survey studies are in the main inconsistent with the incidence study results with larger
effects noted for American samples for depression, paranoia and the two world assumptions
of benevolence and meaningfulness and larger effects for Israeli samples only for anxiety
(see Table 8.11).
The results of Eitinger’s (1972) examination of anxiety and depression levels
among groups of survivors in Israel and Norway afford the first of two direct assessments
of post-war settlement location found by the current author.
The data reveal a 34%
incidence of periods of depression among the 328 Norwegian survivors, with 31% of the
554 Israeli survivors evidencing the same symptoms. Marked symptoms of anxiety were
noted in 36% of the Norwegian sample and 39% of the Israeli sample. Neither of these
differences are statistically significant when analysed by the current author.
© Janine Lurie-Beck 2007
147
The recent study by Kahana, Harel and Kahana (2005) provides the second direct
assessment of post-war settlement location. In this study an American group and an Israeli
group of Holocaust survivors completed the SCL-90 thus providing measures of anxiety,
depression and paranoia.
The American survivors (n = 150) scored statistically
significantly higher than the Israeli survivors (n = 150) on all three variables (Anxiety –
American M = 18.69, SD = 8.61 versus Israeli M = 16.67, SD = 6.98, t (298) = 2.22, p <
0.05; Depression – American M = 25.82, SD = 10.67 versus Israeli M = 21.71, SD = 8.94,
t(298) = 3.60, p < 0.001; Paranoia – American M = 9.45, SD = 3.24 versus Israeli M =
8.64, SD = 3.70, t (298) = 2.01, p < 0.05). The findings here in relation to depression and
paranoia are consistent with the meta-analysis of survey study subsets in Table 8.11 with
American survivors worse off than Israeli survivors. However while this pattern is again
mirrored for anxiety in Kahana, Harel and Kahana (2005), the opposite result is obtained in
the meta-analyses (with Israeli survivors appearing to be worse off than American
survivors).
The findings of Nadler and Ben-Shushan (1989) suggest that the depression levels
of survivors in Israel depends greatly on the nature of the community in which they live.
For example, they found that survivors living within a kibbutz (a very close knit
community group) had statistically significantly lower levels of anxious and low energy
depression than survivors living in the city. No specific statistics can be quoted as they
were not presented by the authors.
Another potentially important influence is the degree to which survivors
successfully integrated into the society and community of the country that they moved to
after the war. This issue was addressed in a study by Hafner (1968). Hafner (1968)
compared a group of survivors whom he considered had integrated well into their new
homeland (“good integration group”) to a group who appeared to have assimilated and
established themselves less well (“bad integration group”).
Among the good social
integration group (n = 102), approximately 18% were found to suffer from recurrent
depression, 32% from chronic depression and 33% from free floating anxiety. Of those
considered to have bad/poor social integration (n = 84), approximately 23% were
diagnosed with recurrent depression, 49% were diagnosed with chronic depression and
40% with free floating anxiety.
Clearly the issue of post-war settlement location warrants further direct
investigation involving direct comparison of survivors stratified by their post-war
© Janine Lurie-Beck 2007
148
settlement location. Issues such as the nature of the community within a country that
survivors settled in as well as the success of their integration into that country need to be
considered. Results from meta-analytic sub-set analysis are inconsistent and there are only
two studies that have considered post-war settlement directly.
8.9. – Membership of Survivor Organisations/Support Groups
Differences in survivor psychological health that relate to membership of a survivor support
group or organisation are examined via sub-set meta-analyses. Studies based on samples
derived from such groups are combined in a separate meta-analysis to those whose samples
were derived from the general community by some other means. Table 8.13 presents the
results of this exercise (see Appendix A for study details).
Apart from the results for paranoia, larger meta-analytic effect sizes are derived
from the studies using a survivor group sample as opposed to the studies using a nonsurvivor group sample. This suggests that survivors who are members of survivor groups
have higher levels of negative symptoms than survivors who are not members of such
groups. This finding is consistent with the argument that the more affected survivors are
more likely to join such groups than less affected survivors (see Chapter Two, Section
2.5.3).
Table 8.13.
Summary of sub-set meta-analyses of survey studies by sample source for survivors
Anxiety
Depression
Paranoia
Positive
Attachment/Intimacy
Variables
Negative
Attachment/Intimacy
Variables
Number of
results whose
survivor
sample was
recruited from
survivor
groups
3
4
2
2
2
Survivor Groups
Survivor
g
group
sample
size
95%
confidence
limits for g
103
123
51
91
0.63 *
0.46 *
0.39 *
-0.38 *
0.36 to 0.89
0.20 to 0.72
0.04 to 0.75
- 0.68 to - 0.09
Number of
results whose
survivor
sample was
recruited from
the general
population
9
24
9
1
91
0.34 *
0.04 to 0.64
1
General Community
Survivor
g
group
sample
size
95% confidence
limits for g
771
1,734
576
39
0.56 *
0.33 *
0.44 *
0.00 #
0.45 to 0.67
0.25 to 0.41
0.32 to 0.56
Cannot compute
39
0.14
- 0.30 to 0.57
Note. An effect size of zero is reported for the study assessing positive attachment dimensions within the general
community because no descriptive data was reported because no statistically significant difference between the survivor
and control groups were found. However, despite the fact that it is highly unlikely that both groups scored exactly the
same, the only unbiased estimate of effect size that can be made in this circumstance is zero.
* p < 0.05
8.10. – Summary and Conclusions
This chapter has presented the findings of meta-analyses and reviews of the literature
addressing demographic differences among Holocaust survivors on a number of
© Janine Lurie-Beck 2007
149
psychological variables. While the results pertaining to gender are relatively consistent, the
data relating to other variables is ambiguous or lacking. While 20 studies were located that
assessed types of Holocaust experiences, there was hardly a consensus on its
operationalisation. Given how much conjecture and anecdotal evidence there is regarding
the influence of the survivors’ loss of family during the Holocaust, with many suggesting it
is the most traumatising element, it is a shame that this has not translated into a more
substantial body of empirical assessment.
Existing data pertaining to the influence of gender, age during the Holocaust, time
lapse since the Holocaust and post-war settlement location were supplemented with metaanalyses by the current author. While these analyses did not always help reach a definitive
conclusion about the role of these variables, it should be noted that these analyses add to
the knowledge already gleaned from the existing data.
© Janine Lurie-Beck 2007
150
Chapter Nine – Meta-Analyses of the Moderating Influence of Demographic Variables
among Descendants of Holocaust Survivors
In Chapter Eight, the influence of numerous demographic variables on the psychological
health of Holocaust survivors was discussed. In this chapter, the influence of demographic
variables on the psychological health of two generations of descendants of Holocaust
survivors is discussed. Both aspects of their survivor parent/grandparents’ experiences
during the Holocaust as well as demographic variables that relate to the children and
grandchildren generations are examined.
9.1. – Method
Direct and indirect assessment of demographic variables and their impact on Holocaust
survivor descendants’ psychological health are presented in this chapter.
Direct
assessments of demographic variables in the literature are summarised and are metaanalysed where possible. Indirect assessment is also possible for some variables.
Demographic variables are assessed indirectly via meta-analytic techniques such as
sub-set meta-analyses and meta-regressions. Sub-set meta-analyses provide insight into
demographic variables such as post-war settlement location via meta-analyses of studies in
sets determined by the country of study.
Meta-regressions are also used to explore
relationships between linear/continuous variables and study effect sizes. Examples of such
continuous variables within the descendants of survivor groups are post war delay in birth,
and the proportion within samples that meet particular criteria (such as the percentage of a
sample that is female or that has two Holocaust survivor parents compared to only one).
By correlating study effect sizes with methodological aspects such as sample criteria,
further analysis of demographic variables is possible using the data already available in the
literature. This is one of the key advantages of a meta-analytic approach to literature
review.
9.2. – Demographic Differences within the Children of Holocaust Survivor/s Group
In this section, the current status of research evidence pertaining to the moderating
influence of demographic variables on the psychological health of children of Holocaust
survivors is established. Demographic variables considered include those relating to their
Holocaust survivor parents (such as the nature of their Holocaust experiences or their age
during the Holocaust) as well as variables that are intrinsic to the children themselves such
as their gender and birth order.
© Janine Lurie-Beck 2007
151
9.2.1. – Number of Survivor parents
Whether a person is the child of one or two Holocaust survivors has the potential to affect
the severity of symptoms they experience, or the degree of trauma transference. The
impact of number of survivor parents is assessed in a number of ways. Results that directly
compared children with one versus two survivor parents are combined meta-analytically
(see Appendix E for study details). In addition to this direct method, studies comparing
children of survivors to control groups are analysed according to the percentage of the
children of survivor sample with two survivor parents (see Appendix B for study details).
This approach is identical to that used to look at the female percentage of study samples for
Holocaust survivors in the previous chapter and children of survivors in this chapter. A
comparison of children of one survivor versus controls and children of two survivors versus
controls is also provided. The results of these analyses are presented in Tables 9.1 and 9.2.
Table 9.1.
Meta-analysis of survey study results based on number of survivor parents
Variable
Depression
Anxiety
Paranoia
Negative
Attachment
Dimensions
Effect size for children with one versus two survivor
parents
Number of results
95%
Total
g
where children
confidence
N
with two survivor
limits for
parents scored
g
higher than with
one survivor
parent
4/5
- 0.19
- 0.43 to
257
0.06
2/4
- 0.01
- 0.29 to
191
0.27
1/4
0.09
- 0.19 to
191
0.37
1/1
- 0.06
- 0.46 to
98
0.36
Effect size for one survivor parent
versus control results
95%
Child of
g
confidence
Survivor
limits for
Group
Sample
g
Size
-0.01
0.03
0.49 *
-0.08
- 0.30 to
0.28
- 0.27 to
0.32
0.19 to
0.79
- 0.39 to
0.24
Effect size for two survivor parents
versus control results
95%
Child of
g
confidence
Survivor
limits for
Group
Sample
g
Size
74
0.10
74
0.05
74
0.38 *
36
-0.01
- 0.16 to
0.35
- 0.23 to
0.34
0.09 to
0.67
- 0.29 to
0.27
117
117
117
62
Notes. A negative effect size indicates that children with two survivor parents scored higher than children with only one
survivor parent.
* p < 0.05
Table 9.2.
Meta-regression of children of survivors versus control results with the percentage of the children of survivor sample with
two survivor parents
Variable
Depression
Anxiety
Paranoia
Positive Attachment Dimensions
Negative Attachment Dimensions
Number of results for which
parental breakdown was
provided and could be included
in meta-regression
13/26
10/17
5/12
4/8
11/13
rmeta
2
rmeta
Child of survivor
sample size
0.02
-0.09
-0.30 **
0.12
0.22 ***
0.00
0.01
0.09
0.01
0.05
547
382
128
247
595
Note. Study results included in meta-regressions are weighted in the analyses by their associated child of survivor
sample size so that results based on larger samples are weighted more heavily to reflect their increased precision in
estimating the true population effect.
** p < 0.01, *** p < 0.001
© Janine Lurie-Beck 2007
152
The results of the meta-analyses are not as strong as anticipated. None of the metaanalyses of results that directly compared children of one to children of two survivor
parents reach statistical significance.
In addition the meta-analytic effect sizes for
children with one survivor parent versus controls are not uniformly weaker than the metaanalytic effect sizes for children with two survivor parents versus controls as would be
intuitively predicted. The results for depression are the only ones that are in the pattern that
would be predicted though none are large enough to reach significance.
It is interesting to note that the results for paranoia uniformly suggest children with
one survivor are more paranoid than children with two survivor parents. This is evidenced
by the larger effect size for children with one survivor parent versus controls than the effect
size for children with two survivor parents versus controls. In addition the meta-regression
which correlates children of survivor versus control effect sizes with the percentage of the
child of survivor sample with two survivor parents is negative suggesting that effect sizes
are lower when the proportion of the sample with two survivor parents is higher. The
children with one survivor parent versus controls meta-analytic effect size, children with
two survivor parents versus controls meta-analytic effect size and the meta-regression with
the percentage of the child of survivor sample with two survivor parents are all statistically
significant for paranoia. The only other statistically significant result from these analyses is
the statistically significant positive correlation between the percentage of the child of
survivor sample with two survivor parents and effect sizes comparing children of survivors
to controls on negative attachment dimensions.
This result suggests that children of
survivors score much higher than controls on negative attachment dimensions when they
have two survivor parents than if they only have one survivor parent.
One study had to be left out of meta-analytic calculations due to sample
contamination. Rubenstein (1981) also attempted to examine differences in depression
between children of one survivor and children of two survivors. This study included
subjects born before 1945 which, for the purposes of this study, are not considered to be
members of the children of survivor population but rather are survivors themselves. It is
worthy of note that, despite the problems with Rubenstein’s (1981) study, his findings do
add further support to the idea that children of two survivor parents show higher levels or
greater severity of psychological symptoms. Using the Depression scale of the Mini-Mult
(an abbreviated version of the MMPI), Rubenstein’s (1981) study found that children with
two survivor parents had higher depression scores than children with one survivor parent
© Janine Lurie-Beck 2007
153
(M = 22.40, SD = 4.88, n = 48 and M = 20.75, SD = 4.52, n = 30 respectively). However,
this difference did not reach statistical significance.
9.2.2. – Gender of Survivor Parent
Despite the many opportunities to examine the influence of survivor parent gender in
studies with children of survivors (particularly those that delineated children with one and
two survivor parents) only one study was located that directly assessed this variable.
Schleuderer (1990) not only noted whether participants had one or two Holocaust survivor
parents, but, unlike other researchers who obviously had the same opportunity, he also
noted the gender of the Holocaust survivor parent for those with only one survivor parent.
On both measures of depression used in his study (the Dysthymic and Major Depressive
subscales of the Millon Clinical Multiaxial Inventory II), children with Holocaust survivor
fathers had higher levels of depression than those whose mother was the survivor. On the
Dysthymic subscale, children of survivor fathers were given a rating of 9.84 (SD = 11.39, n
= 25) while children of survivor mothers were given a rating of 7.92 (SD = 9.61, n =12).
On the Major Depression subscale, children of survivor fathers were rated at 6.64 (SD =
8.26) and children of survivor mothers were rated at 5.50 (SD = 6.17). Children with
survivor fathers also had the higher levels of anxiety (M = 5.48, SD = 7.17 versus M = 4.33,
SD = 4.29) and paranoia (M = 25.20, SD = 12.23 versus M = 24.08, SD = 15.12) compared
to those with survivor mothers. None of these differences reached statistical significance
but it is interesting to note the consistency of results with children of survivor fathers
scoring less favourably on all measures.
9.2.3. – Type of Survivor parent’s Holocaust experiences
Only five studies were found to have considered the impact of survivor parents’ type of
Holocaust experiences on their children. Because of the different categories and analysis
methods used by these studies, meta-analytic synthesis of the results is not possible.
However each study is discussed in turn in this section.
In the first study, Lichtman (1983; 1984) divided her sample into children with one
parent who had spent time in a concentration camp, two parents who had been in a camp,
one or two parents who had been in hiding and one or two parents who had escaped. As
measured by the MMPI depression subscale, children with one parent who had been in a
camp had the lowest depression scores (M = 20.58, SD = 7.20, n = 21), followed by
children of two camp survivors (M = 21.01, SD = 4.46, n = 21), children of one or two
survivors who escaped (M = 22.10, SD = 5.05, n = 13) and children of one or two survivors
© Janine Lurie-Beck 2007
154
who were in hiding (M = 23.10, SD = 7.52, n = 9). On the MMPI anxiety subscale,
children with both parents as camp survivors had the lowest anxiety (M = 11.21, SD = 7.87,
n = 21), with increasing levels of anxiety among children of one camp survivor (M = 13.46,
SD = 8.60, n =21), children of one or two survivors who escaped (M = 14.89, SD = 6.55, n
= 13) and children of one or two survivors who were in hiding (M = 20.96, SD = 13.30, n =
9). On the MMPI paranoia subscale, children of one or two survivors who escaped (M =
8.46, SD = 3.33, n = 13) had the lowest paranoia, followed by children with both parents as
camp survivors (M = 9.67, SD = 3.50, n = 21), children of one camp survivor (M = 9.81, SD
= 4.17, n = 21) and children of one or two survivors who were in hiding (M = 11.89, SD =
4.01, n = 9). None of the differences between these subgroups were statistically significant.
It is interesting to note that for both anxiety and depression it was the children of camp
survivors who had the lowest levels of symptoms compared to children of survivors who
were in hiding and survivors who escaped at some point. The pattern is not exactly
replicated for paranoia but it is notable that children of survivors who were in hiding were
also the ones with the highest paranoia levels. However, it is also duly noted that these
results were based on very small sample sizes and this paired with the lack of significance
means that no firm assertions can be made based on this study.
Lichtman (1983; 1984) also examined the impact of parental Holocaust experience
on children of survivors in another less direct way. Subjects were asked to rate the severity
of their survivor parent’s Holocaust experiences (as they perceived them).
Factors
incorporated into this questionnaire included descriptions of parental war experiences on a
scale from fortunate to catastrophic, an assessment of the effect the Nazi occupation had on
parental lifestyle from enhancing to devastating and the nature of conditions lived under
during the war from pleasant to unbearable. The correlation between rating of parental
trauma and depression among children with two survivor parents was 0.10 (n = 52). The
correlation between rating of parental trauma and anxiety among children with two survivor
parents was 0.07 (n = 52). Neither of these correlations were statistically significant. The
correlation between rating of parental trauma and paranoia among children with two
survivor parents was 0.26 (n = 52). This correlation narrowly missed the statistically
significant cut off (p = 0.06). These results are not exactly earth shattering either and the
measure of parental trauma is not pure as it includes perceptions of the children and is not
objectively based on the details of the survivor parents experiences but on the perception of
the impact of those experiences.
© Janine Lurie-Beck 2007
155
In the second of the five studies to be discussed in this section, Leon, Butcher,
Kleinman, Goldberg and Almagor (1981) looked at how different parental Holocaust
experiences influences children’s depression and paranoia levels. They separated their
group into children of survivors who had been interned in a concentration camp and those
who had other Holocaust experiences. Interestingly there was a differential effect by
gender with daughters of survivors with camp experience more depressed than those of
survivors with other experience (M = 56.69, SD = 9.67, n = 16 versus M = 52.60, SD =
8.32, n = 5), but the opposite effect for sons (M = 59.93, SD = 12.16, n = 16 for sons of
camp survivors versus M = 62.25, SD = 16.76, n = 4 for sons of non-camp survivors). Leon
et al. (1981) also assessed how different Holocaust experiences of parents may affect
children’s paranoia levels. A slightly different interaction effect was noted for paranoia
with female children of camp survivors and non-camp survivors roughly equivalent in their
paranoia levels (M = 53.75, SD = 8.23, n = 16 and M = 54.20, SD = 8.90, n = 5
respectively) and a non-statistically significant but much more marked difference among
male children with those with camp survivor parents more effected (M = 58.06, SD = 7.69,
n = 16 versus M = 51.25, SD = 5.31, n = 4). As with Lichtman’s (1983; 1984) findings,
these results were based on very small sample sizes and therefore should be interpreted
with caution.
In the third study, using a crude measurement of differential Holocaust
traumatisation, Schleuderer (1990) conducted a correlation between subject’s rating of
parental Holocaust trauma and measures of psychological symptoms. Participants gave a
rating between 1 (minimal) and 5 (extreme) regarding their impression of the traumatic
level of their mother and father’s Holocaust experiences. The correlations between their
rating of paternal trauma (n = 100) and dysthymia was -0.02, major depression 0.06,
anxiety 0.12, and paranoia – 0.11. For maternal trauma (n = 100) the correlations were:
dysthymia 0.04 (n = 100), major depression 0.01, anxiety 0.17, and paranoia 0.04. None
of these correlations were statistically significant and all are of negligible strength.
In the fourth study, Brom, Kfir and Dasberg (2001) calculated a severity of
Holocaust experience score for the parents of their daughters of survivors group. The score
was derived by the research team by taking into account the survivor parents’ experiences
during the war such as the kind of persecution they experienced, their age and the number
of family members they lost.
This severity of Holocaust experience score was then
correlated with the daughters’ depression scores.
© Janine Lurie-Beck 2007
The correlation between these two
156
variables was 0.23 (n = 31) which was not statistically significant. Again this measure of
parental trauma is not ideal and there is no indication of how the survivor’s age was taken
into account in the trauma measure – was it assumed that age was positively or negatively
associated with severity of symptoms?
In the last study located that considered the impact of the type of Holocaust
experiences endured by a survivor on the psychological health of their children, Bauman
(2003) looked at children with survivor mothers and fathers separately depending on
whether their survivor parent had been in a camp or had other Holocaust experiences.
Participants completed the neuroticism subscale of the NEO Five Factor Inventory which
measures predisposition to anxiety, hostility, depression, self-consciousness, impulsivity
and vulnerability. Among the children of Holocaust survivor mothers there was very little
difference relating to whether the survivor mother had been in a camp (M = 3.60, SD =
1.23, n = 68) or had other experiences (M = 3.57, SD = 1.25, n =82). There was a larger
gap between the scores of children with a Holocaust survivor father who had been in a
camp (M = 3.46, SD = 1.24, n = 40) compared to children whose father had other
Holocaust experiences (M = 3.60, SD = 1.21, n = 108), however neither difference reached
statistical significance. It is interesting to note though that the results relating to survivor
father here seem to mirror those of Lichtman (1983; 1984) in that the children of non-camp
survivors score higher than children of camp survivors.
While there is some tentative evidence here to suggest that children of non-camp
survivors may evidence higher levels or severity of symptoms it is by no means
indisputable.
For the most part, this issue has been examined with inadequate
operationalisation or small samples. Given the strong evidence for differential impact
based on type of experience on survivors themselves it is reasonable to assume that there
are discernible differences among descendants that are dependent on the Holocaust
narrative of their ancestors.
However, the exact translation to the child of survivor
generation has not been clearly delineated to date.
9.2.4. – Parental Loss of Family Members
Theoretical discourse and other empirical findings discussed in earlier chapters would
suggest that children of survivors who lost family members during the Holocaust would be
more adversely affected in psychological functioning than children of survivors whose
families stayed in tact. Only two studies were located that considered the impact of a
survivor’s loss of family members on the subsequent psychological health or functioning of
© Janine Lurie-Beck 2007
157
their children. Both point to more negative affects among children of survivors being
associated with more familial loss on the part of their survivor parents.
Schwarz (1986) gathered data on the number of children that female survivors lost
during the war and levels of phobic anxiety among children born to the female survivors in
the post-war period. The correlation co-efficient derived was 0.40 (n = 67) which was
statistically significant (p < 0.001). In other words, the more children lost by a female
survivor during the war, the higher the anxiety level experienced by her children born in the
post-war period.
Gertler (1986) measured the degree to which achieving intimacy is difficult among
children of survivors who lost a spouse or a child/children during the war compared to
children of survivors who did not lose a spouse or child during the war. While this
difference did not reach statistical significance the children of survivors who lost family
members scored higher on the Hard to be Intimate scale than children of survivors whose
families escaped death (M = 1.01, SD = 0.72, n = 26 versus M = 0.86, SD = 0.55, n = 72).
There is a consistency between these two findings that suggest greater familial
losses by Holocaust survivors may well be associated with more negative outcomes for
their children. However, two studies provides inadequate assessment of this issue and
further data is required here.
9.2.5. – Survivor Parent/s Country of Origin
Two studies were located that analysed the level/severity of symptoms among children of
survivors depending on their survivor parents’ country of origin. As outlined in Chapter
Four, a survivor’s country of origin has implications for the nature and duration of their
persecution during the Holocaust which in turn relates to the degree of traumatisation they
evidenced in the post-war period. This variable therefore also has the potential to impact
the severity of transmission of trauma to the next generation.
Schleuderer (1990) cross-referenced his children of survivor sample with the
country of origin of their survivor mothers and/or fathers. While sample sizes were quite
small with all but one of the nine countries listed having samples less than 10, there were
notable differences in dysthymia, major depression, anxiety and paranoia scores. Listing all
the means and standard deviations for all the groups would be tedious and largely
meaningless given the sample sizes and lack of statistical significance but they certainly
pointed to potential statistically significant effects if larger sample sizes had been used.
© Janine Lurie-Beck 2007
158
Berger (2003) examined the impact of survivor parents country of origin on their
children’s romantic attachment style. However, because she included children of nonsurvivors in the analysis it is not clear whether some of the parents born in European
countries were not survivors. In addition, she did not specify the sample sizes of each of
the country sub-groups. Statistically significant differences were obtained on both the
avoidance and attachment anxiety dimensions when considering mother’s country of origin
and a statistically significant result was also obtained when considering the impact of
father’s country of origin on avoidance. In relation to mother’s country of birth, children of
Romanian-born mothers had the highest scores on avoidance (M = 74.86), followed by
German (M = 57.11), American (M = 48.25), Polish (M = 44.13) and finally Hungarian (M
= 39.22). According to ANOVA analysis, these differences are statistically significant (p <
0.01).
Interestingly, some marked differences in rank were noted when looking at
attachment anxiety, particularly for children of Hungarian-born mothers: Romanian (M =
80.29), Hungarian (M = 79.33), American (M = 73.98), German (M = 57.83), and finally
Polish (M = 56.05). These differences also reached significance in an ANOVA analysis (p
< 0.05). Turning to father’s country of origin, the results for avoidance followed a similar
pattern to that found in relation to mother’s country of origin.
Again, children of
Romanian-born father’s were highest on the list (M = 71.00), followed by German (M =
53.43), Polish (M = 49.46), American (M = 47.17), Czech (M = 40.30) and finally
Hungarian (M = 31.56).
These results were statistically significantly different when
analysed via ANOVA (p < 0.05). On the attachment anxiety dimension, children of
Romanian fathers were rated the highest (M = 74.40), followed by Hungarian (M = 66.44),
American (M = 65.00), German (M = 64.79), Czech (M = 63.25) and Polish (M = 61.74).
None of these differences reached statistical significance.
While this evidence is by no means overwhelming, it is suggestive of the reasoning
that children of survivors’ may have experienced differential rates of trauma transmission
depending on their survivor parents’ country of origin. The influence of survivor’s country
of origin has been inadequately assessed for both the survivor and child of survivor
generations.
9.2.6. – Age of Survivor Parent/s during the Holocaust.
Four studies directly assessed the impact of survivors’ age during the Holocaust on the
psychological health and/or functioning of their children. Because the age categories used
© Janine Lurie-Beck 2007
159
by each study differed widely it was not possible to combine these results metaanalytically.
Budick (1985) compared children of survivors who were teenagers between 1939
and 1945 (aged 13 to 18) and those who were adults (aged 19 and over). The children of
the older survivors had the higher level of depression, as measured by the depression
subscale of the Basic Personality Inventory, (M = 3.56, SD = 3.22, n = 16 versus M = 2.50,
SD = 2.25, n = 16). This pattern was matched on the anxiety subscale of the Jackson
Personality Inventory (M = 10.93, SD = 4.01 versus M = 10.37, SD = 3.12); Interpersonal
Affect (M = 13.18, SD = 3.89, n = 16 versus M = 12.31, SD = 4.73, n = 16) and Succorance
(M = 9.25, SD = 4.78, n = 16 versus M = 7.88, SD = 3.84, n = 16). None of these
differences reached statistical significance but that is unsurprising given the sample sizes
involved. Budick (1985) controlled for the possible confounding variable of the time lapse
between the children’s birth and the end of the war. Both groups of children had a similar
age range (children of teenage survivors ranging from 21 to 36 and children of adult
survivors ranging from 24 to 37).
In the second study, Eskin (1996) calculated a correlation between children of
survivors’ depression scores and their parent’s age. This provided another indirect measure
of the impact of survivors’ age during the Holocaust and depression in their children. The
correlation between these two variables was -0.18 (n = 49) and was not statistically
significant.
However, although quite weak this result suggested that in this sample of
children of survivors, depression was higher the younger their parents were (and therefore
the younger they must have been during the Holocaust).
Thirdly, Berger (2003) analysed children of survivors’ scores on attachment-related
avoidance and anxiety in relation to their mother and father’s age during the Holocaust.
Survivor parents’ age during the Holocaust was operationalised categorically as “child”,
“adolescent” and “adult”. No age definitions were provided for these categories. Another
problem is that some mothers and fathers were not survivors but it is not clearly stated that
children with non-survivor parents were removed from the analyses. Therefore it is not at
all certain if this is a pure assessment of this variable; however in the interests of
completeness, it is interesting to consider the results even if very cautiously. In relation to
mother’s age first, children whose mothers were adolescents during the war scored the
highest on avoidance (M = 52.76, SD = 22.42), followed by those whose mothers were
adults (M = 49.37, SD = 23.72) and finally those whose mothers were children (M = 45.31,
© Janine Lurie-Beck 2007
160
SD = 23.84). No statistically significant difference was found. In terms of anxiety a more
linear (though again, non significant) relationship was found with children of adults
mothers most anxious (M = 69.23, SD = 23.21), followed by adolescents (M = 63.32, SD =
27.70) and children (M = 60.86, SD = 26.83). The impact of father’s age during the war on
avoidance was similar to that of mothers in pattern and non-significance (child M = 47.17,
SD = 26.66, adolescent M = 45.08, SD = 22.81, adult M = 50.51, SD = 23.05). The children
of fathers who were children during the war scored the highest on attachment anxiety (M =
74.75, SD = 32.40) followed by those whose fathers were adults (M = 65.45, SD = 26.09)
and finally adolescents (M = 58.00, SD = 23.45).
Finally, Bauman (2003) compared children of survivors scores on neuroticism
depending on whether their survivor mother or father was a child or adolescent during the
war or if they were an adult. Again, similar to Berger (2003) the age ranges of these
categories was not provided.
Children of survivor mothers who were children or
adolescents during the war scored higher on neuroticism than children of survivor mothers
who were adults during the war (M = 3.65, SD = 1.27, n = 95 versus M = 3.34, SD = 1.10,
n = 55). A much smaller and negligible difference was noted among children of survivor
fathers (Child/Adolescent M = 3.47, SD = 1.19, n = 38 versus Adult M = 3.55, SD = 1.20,
n = 110). Neither of these differences reached significance.
There are mixed findings here with some results suggesting children of survivors’
symptoms increasing with their survivor parents’ age during the war and others suggesting
they decrease. It is noted though that the one study finding children of older survivors to be
more affected had very small samples. This section has highlighted yet another variable
that requires further clarification.
9.2.7. – Length of time between the end of the war and the birth of children.
This issue has never been addressed directly in a study with children of Holocaust
survivors. Three studies were located that consider the impact of a child of survivor’s age
which provides an indirect measure of this issue (E. Berger, 2003; Eskin, 1996; Gertler,
1986). The results of these three studies are summarised first and then meta-regressions
conducted by the current author are presented.
The correlation presented by Eskin (1996) between depression among children of
survivors and their age provided an indirect measure of this relationship for depression.
The older they children of survivors are the shorter the time delay between the end of the
war and their birth. The correlation was - 0.11 (n = 49). While not statistically significant
© Janine Lurie-Beck 2007
161
this suggested a weak negative relationship where the older the subjects were the less
depressed they were.
In the second study, Gertler (1986) allowed an examination of this issue by dividing
his children of survivor group into age groups based on their age at the time of the study.
These age group boundaries can be used to determine a range in the length of time between
their birth and the end of the war. The three age groups were 25 to 30, 31 to 35 and 36 to
40 in 1986. So the group aged between 36 to 40 were born between 1946 and 1950, those
aged 31 to 35 were born between 1951 and 1955 and those aged 25 and 30 were born
between 1956 and 1961. These correspond to a post-war delay in birth of 1 to 5 years, 6 to
10 years and 11 to 15 years respectively. As was mentioned before, these groups were also
segmented by number of survivor parents. In both array cases, it was the children born
with the shortest time delay after the war who scored the highest on the Hard to be Intimate
scale. No statistically significant differences were found but this is not surprising given the
small cell numbers. Among the children with one survivor parent, children born with a 1 to
5 year delay scored 1.11 (SD = 1.01, n =3), a 6 to 10 year delay scored 0.78 (SD = 0.79, n =
18) and 11 to 15 year delay 0.79 (SD = 0.54, n = 15). Among the children with two
survivor parents, the 1 to 5 year delay group scored 1.07 (SD = 0.75, n = 25), a 6 to 10 year
delay scored 0.86 (SD = 0.53, n = 24) and 11 to 15 year delay 0.92 (SD = 0.49, n = 13).
Finally, the correlation between children of survivor’s age and their scores on
attachment related avoidance and anxiety presented in Berger (2003) provides another
indirect measure of the impact of post-war delay in birth. She obtained a correlation of
0.11 between children of survivor’s age and avoidance and 0.01 between age and anxiety.
This correlation is no stronger that that found by Eskin (1996) and the two results together
suggest that perhaps post-war delay in birth is not a very influential variable.
Meta-regressions have been conducted by the current author between children of
survivors versus control results and the average delay in between 1945 and the children of
survivor sample’s birth. Average delay in birth is calculated using the average age of the
children of survivor samples as cited by researchers and the study year.
Working
backwards, the average year of birth is determined and from this the average delay between
1945 and the average year of birth. The results of these analyses are presented in Table 9.3.
Details of the studies included in this meta-regression can be found in Appendix B.
© Janine Lurie-Beck 2007
162
Table 9.3.
Meta-regression of children of survivor versus control results with average delay between child of survivor birth and 1945
Variable
Depression
Anxiety
Paranoia
Positive Attachment Dimensions
Negative Attachment Dimensions
Number of results with age
data provided which could be
included in the metaregression
17/26
14/17
7/12
2/8
3/13
rmeta
2
rmeta
-0.47 ***
-0.38 ***
0.53 ***
1.00 ***
0.24 **
0.22
0.15
0.28
1.00
0.06
Child of survivor sample size
774
492
213
70
124
Note. Study results included in meta-regressions are weighted in the analyses by their associated child of survivor
sample size so that results based on larger samples are weighted more heavily to reflect their increased precision in
estimating the true population effect.
** p < 0.01, *** p < 0.001
The meta-regressions for depression and anxiety are in the direction that is
consistent with theoretical conjecture in the literature. Both of these analyses suggest
higher levels of depression and anxiety (or greater disparity between children of survivors
and control groups) among children of survivors with a shorter delay between the end of
the war and their birth. The result for paranoia is in the opposite direction, suggesting
higher paranoia with longer post-war delays in birth.
The results for negative and positive attachment dimensions are more ambiguous.
The perfect correlation of 1.00 for positive attachment dimensions is in the desirable
direction but is clearly a statistical aberration given the small number of results involved in
the analysis. The meta-regression for negative attachment dimensions repeats the pattern
found for paranoia in that it is in the opposite direction to what would be predicted.
However, again there are sample size and result number issues here and so very little
meaning can be attached to this result.
9.2.8. – Location of post-war settlement.
Differences in psychological health and functioning among children of survivors that relate
to their survivor parent/s post-war settlement location have only been directly assessed by
one study (Okner & Flaherty, 1988). However this issue can also be examined indirectly
via a meta-analytic comparison of studies conducted in different countries/regions.
Okner and Flaherty (1988) compared children of Holocaust survivors who had
grown up in Israel to those who had grown up in America on measures of depression and
anxiety. While there were no statistically significant differences between the two groups, it
was the American group who scored higher on both the depression and anxiety measures
than their Israeli counterparts (Depression M = 12.60, SD = 11.30, n = 138 versus M =
11.90, SD = 10.10, n = 52; Anxiety M = 13.30, SD = 8.00 versus M = 12.60, SD = 8.10)
© Janine Lurie-Beck 2007
163
Two meta-analytic exercises have been carried out by the current author with postwar settlement location.
Sub-set meta-analyses for survey studies and sub-set meta-
analyses for incidence studies provide an additional assessment of the influence of this
variable and are reported in Tables 9.4 and 9.5. Studies included in these sub-set metaanalyses can be found in Appendix B.
Table 9.4.
Summary of sub-set meta-analyses of survey studies by post-war settlement location for children of survivors
Number of results
based on data
collected in Israel
Israel
Child of
survivor
sample size
g
95%
confidence
limits for g
Anxiety
2
95
0.00
Depression
Positive
Attachment
Dimensions
2
1
75
70
0.15
- 0.37 *
Not able to
compute
- 0.17 to 0.47
- 0.73 to - 0.00
Number of
results based on
data collected in
America
15
24
7
America
Child of
survivor
sample size
g
95%
confidence
limits for g
518
0.21 *
0.09 to 0.34
1,012
348
0.10 *
- 0.05
0.00 to 0.19
- 0.19 to 0.09
Notes. Both results for anxiety stated there were no statistically significant differences but did not quote descriptive data.
Therefore a null result (or mean difference of zero) was entered into the analysis for both studies. It is highly likely that
the groups did differ in some direction, though not statistically significantly and it cannot be determined what the true
effect size would have been.
* p < 0.05
Table 9.5.
Summary of sub-set meta-analyses of incidence/diagnosis studies by post-war settlement location for children of
survivors
Variable
Number of
studies
conducted
in America
Ω
Anxiety
3
3.16 *
Depression
3
6.96 *
America
95% confidence
Average
Incidence
Child of
Survivor
Sample
Size
2.07 to 4.82
19%
4.30 to 11.26
23%
limits for
Ω
340
Number
of studies
conducte
d in
Europe
1
3.19
340
1
3.64
Ω
Europe
95%
confidence
limits for
Average
Incidence
Child of
Survivor
Sample
Size
53%%
59
34%
59
Ω
0.30 to
33.89
0.78 to
16.93
Variable
Israel
Number of
95%
Average
Child of
Ω
studies
confidence
Incidence
Survivor
conducted in
Sample
limits for Ω
Israel
Size
Anxiety
1
1.00
0.48 to 2.06
7%
147
Depression
1
0.83
0.26 to 2.65
3%
147
Note. European studies conducted in Norway and France. Both average Europe incidences include the findings of Zajde’s (1998) study conducted in
France. Zajde (1998) did not compare her results to a control group and so her data could not be included in the odds
ratio analysis.
Ω meta-analytic odds ratio
* p < 0.05
From a perusal of the sub-set meta-analysis of survey results (Table 9.4) it appears
that children of survivors in Israel have faired worse than their counterparts in America.
However, looking at the meta-analysis of incidence studies it is the children of survivors
based in Israel who have much lower incidence of both anxiety and depression (though this
© Janine Lurie-Beck 2007
164
is based on only one study). The survey results meta-analysis finding that Israeli children
of survivors score worse than American children of survivors contradicts the direct
assessment of this issue by Okner and Flaherty (1988), however the incidence study metaanalysis is clearly supportive of their findings.
It seems there is still too much
inconsistency in the available date pertaining to post-war settlement and its influence on
children of survivors to be able to make a firm conclusion.
9.2.9. – Gender
Gender differences in the psychological well-being of children of survivors are analysed in
a similar fashion to analyses conducted for Holocaust survivors in Chapter Eight. Firstly,
results that directly compared male and female survivors are meta-analytically combined in
effect size and odds ratio meta-analyses (see Appendix F for studies included). In addition,
an indirect method is used in which studies comparing survivors to control groups re
analysed according to the female percentage of their sample (see Appendix B).
A
comparison of male versus control and female versus control is also provided (see
Appendix B). The results of these analyses are presented in Tables 9.6, 9.7, 9.8 and 9.9.
Table 9.6.
Meta-analysis of survey study results based on child of survivor gender
Variable
Depression
Anxiety
Paranoia
Positive
Attachment
Dimensions
Negative
Attachment
Dimensions
Effect size for male versus female results
Number of
results where
females
scored higher
than males
12/18
9/12
6/10
2/2
g
95%
confidence
limits for g
Total
N
Effect size for male versus control
male results
95%
Child of
g
confidence
Survivor
N
limits for g
Effect size for Females versus control
female results
95%
Child of
g
confidence
Survivor
N
limits for g
- 0.22 *
- 0.20
0.17
- 0.54
- 0.38 to - 0.05
- 0.41 to 0.02
- 0.06 to 0.40
- 1.25 to 0.17
749
427
324
32
0.03
0.19
0.03
0.19
- 0.13 to 0.20
- 0.07 to 0.45
- 0.27 to 0.33
- 0.51 to 0.89
373
125
77
16
0.17 *
0.21
- 0.03
0.09
0.02 to 0.31
- 0.00 to 0.43
- 0.34 to 0.28
- 0.62 to 0.79
371
180
77
16
3/4
- 0.15
- 0.51 to 0.21
130
- 0.24
- 0.54 to 0.06
84
0.17
- 0.02 to 0.35
261
Note. A negative effect size for males versus females indicates that females scored higher than males.
* p < 0.05.
© Janine Lurie-Beck 2007
165
Table 9.7.
Meta-regression of children of survivor versus control results with the female percentage of the child of survivor sample
Variable
Number of results for
which render breakdown
was provided and could
be included in the metaregression
20/26
10/17
7/12
4/8
5/13
Depression
Anxiety
Paranoia
Positive Attachment Dimensions
Negative Attachment Dimensions
rmeta between
effect size and
female % of
sample
0.26 ***
0.09
0.21 **
-0.13
0.22 **
2
rmeta
Child of survivor
sample size
0.07
0.01
0.04
0.02
0.05
950
428
190
215
259
Note. Study results included in meta-regressions are weighted in the analyses by their associated child of survivor
sample size so that results based on larger samples are weighted more heavily to reflect their increased precision in
estimating the true population effect.
** p < 0.01, *** p < 0.001
Table 9.8.
Summary of meta-analyses of incidence/diagnosis studies comparing male and female children of survivors
Variable
Anxiety
Number of
results with
higher incidence
among females
than males
0/1
Ω
95% confidence
-1.08
0.42 to 2.82
limits for
Ω
Average
incidence
among females
Average incidence
among males
Child of survivor
sample size
27%
29%
93
Ω meta-analytic odds ratio
Table 9.9.
Meta-regression of incidence rates among children of survivors with the female percentage of the child of survivor sample
Variable
Number of results for
which gender
breakdown was provided
and could be included in
the meta-regression
3/5
3/5
Depression
Anxiety
rmeta between
incidence and
female % of
sample
- 0.88 ***
- 0.23 ***
2
rmeta
Child of survivor
sample size
0.77
0.05
233
233
Note. Study results included in meta-regressions are weighted in the analyses by their associated child of survivor
sample size so that results based on larger samples are weighted more heavily to reflect their increased precision in
estimating the true population effect.
*** p < 0.001
Overall, the survey studies seem to suggest that sons of survivors have faired better
than daughters of survivors. However, the incidence studies seem to suggest the opposite.
However, the survey study results are based on a larger number of studies and also a larger
sample of children of survivors and so perhaps more credence can be given to those results.
Certainly the meta-regressions with female percentage of sample are a much cruder and
more indirect way of assessing this issue. The incidence study that directly compared a
male and female sample found very similar incidence levels for the genders. On balance
the studies that directly compared male and female children of survivors, be it with a survey
or a diagnostic or incidence study found that female children of survivors are worse off
compared to male children of survivors.
© Janine Lurie-Beck 2007
166
9.2.10. – Birth Order
Three studies were located that considered children of survivors’ birth order as a
demographic variable that could potentially impact on psychological health (Eskin, 1996;
Lowin, 1983; Schleuderer, 1990). Each used a different mode of operationalisation of birth
order and so meta-analyses could not be conducted.
In Eskin’s (1996) study, participants who were the only child of a survivor or
survivors had the highest depression level on the Centre for Epidemiologic Studies
Depressed Mood Scale (M = 35.30, SD = 11.90, n = 7), followed by middle children (M =
35.00, SD = 11.6, n = 5), first born children (M = 31.30, SD = 9.6, n = 28) and youngest
children (M = 30.10, SD = 4.60, n = 9). No difference reached statistical significance.
Using the SCL-90 depression subscale Lowin’s (1983) second born children had
higher depression levels than their first born counterparts (M = 1.16, SD = 087, n = 23
versus M = 0.81, SD = 0.66, n = 21). A similar pattern was obtained for the anxiety
subscale (M = 0.85, SD = 0.70 versus M = 0.64, SD = 0.57). Neither of these differences
were statistically significant. However, on the paranoia subscale Lowin’s (1983) second
born children had statistically significantly higher paranoia levels than their first born
counterparts (M = 1.19, SD = 0.84, n = 23 versus M = 0.73, SD = 0.58, n = 21; t (42) = 2.06, p = <0.05).
Schleuderer (1990) looked at the levels of dysthymia, major depression, anxiety and
paranoia among children of survivors in five birth order positions. Again contrary to
theories that first born children are the most affected it was the middle order children that
appeared to be most depressed, anxious and paranoid. Mean dysthymia levels for the
groups were as follows: first born (M = 9.77, SD = 11.83, n = 47), second born (M = 11.26,
SD = 13.01, n = 38), third born (M = 9.33, SD = 7.23, n = 12), fourth born (M = 14.00, SD
= 4.24, n = 2), fifth born (2, n = 1). For major depression the statistics were: first born (M =
6.72, SD = 9.09, n = 47), second born (M = 7.66, SD = 9.81, n = 38), third born (M = 7.17,
SD = 6.55, n = 12), fourth born (M = 6.50, SD = 3.54, n = 2), fifth born (2, n = 1). Mean
anxiety levels for the groups were as follows: first born (M = 5.11, SD = 6.69, n = 47),
second born (M = 6.16, SD = 7.62, n = 38), third born (M = 7.17, SD = 7.09, n = 12), fourth
born (M = 4.00, SD = 1.41, n = 2), fifth born (3, n = 1). None of these differences are
statistically significant. The middle order children had the highest paranoia levels (first M
= 22.36, SD = 11.28, n = 44; second M = 22.50, SD = 12.31, n = 38; third M = 26.83, SD =
9.17, n = 12; fourth M = 20.50, SD = 0.71, n = 2; fifth 65.00, n = 1). The only statistically
© Janine Lurie-Beck 2007
167
significant difference found among these groups was for the third born versus the fourth
born with the former group statistically significantly more paranoid than the latter, however
there were only two fourth born children (t (12) = 2.22, p < 0.05).
All three studies suggest that in survivor families with more than one child it is the
middle order children that have higher symptom levels than the first born children which is
contrary to what many would predict (S. Davidson, 1980a; Grubrich-Simitis, 1981;
Newman, 1979; Porter, 1981).
Only one study looked at only children compared to
children from multiple children families.
There are however, as has been the case
frequently in this body of research, quite small sample sizes involved in most cases. The
issue of birth order has been far from adequately researched to date.
9.2.11. – Membership of Descendants of Survivors Organisations or Support Groups
As was conducted for Holocaust survivors, sub-set meta-analyses are calculated for studies
based on samples derived from descendant organisations and those derived from the
general community.
The results of these analyses are presented in Table 9.10 (see
Appendix B for study details). Apart from anxiety, stronger results are obtained for studies
whose samples were obtained from descendant organisations. This suggests that children
of survivors who are members of a descendant organisation are more depressed and
paranoid and experience more negative and less positive attachment attributes, mirroring
the sub-set meta-analysis findings addressing the same issue in the survivor generation.
Table 9.10.
Summary of sub-set meta-analyses of survey studies by sample source for children of survivors
Descendant of Survivor Groups
Number of
Child of
95%
g
results based on
survivor
confidence
samples
sample
limits for g
recruited from
size
descendant of
survivor groups
Anxiety
Depression
Paranoia
Positive
Attachment/Intimacy
Variables
Negative
Attachment/Intimacy
Variables
10
9
4
4
416
366
193
177
0.12
0.20 *
0.40 *
-0.15
- 0.02 to 0.27
0.04 to 0.36
0.16 to 0.64
- 0.37 to 0.07
Number of
results
based on
samples
recruited
from the
general
community
6
17
8
3
7
329
0.11
- 0.05 to 0.26
6
General Community
Child of
g 95% confidence
survivor
limits for g
sample
size
159
721
105
203
0.16
0.05
0.01
-0.13
- 0.04 to 0.37
- 0.06 to 0.16
- 0.24 to 0.26
- 0.31 to 0.05
329
-0.05
- 0.19 to 0.10
* p < 0.05
© Janine Lurie-Beck 2007
168
9.3. – Grandchildren of Holocaust Survivors
Very little research has been conducted with grandchildren of Holocaust survivors to date.
It is granted that it has only been in recent years that grandchildren of survivors have
reached adulthood in large enough numbers to make conducting research with them
feasible.
Therefore very little data exists on demographic differences within the
grandchildren of Holocaust survivor population. What little data there is in this regard is
presented in the following subsections.
9.3.1. – Gender
Jurkowitz (1996) compared male and female grandchildren using the Centre for
Epidemiologic Studies Depressed Mood Scale. It was the females (M = 11.37, SD = 10.01,
n = 47) who scored higher on this scale than males (M = 10.14, SD = 7.44, n = 44),
consistent with other findings that females of differing generations are more depressed than
males. This result was not statistically significant.
9.3.2. – Number/Gender of Child of Survivor Parent/s
Gopen’s (2001) study delineated participants based on whether their mother, father or both
were children of survivors.
Of the measures used by Gopen (2001), the Perceived
Relationships Quality Components Inventory Intimacy and Trust subscales were considered
relevant to the current review. When the grandchildren of survivor sample was stratified
according to the number and gender of children of survivor parents it was the participants
with only a child of survivor father who scored the lowest on both scales (intimacy M =
5.27, SD = 0.43 and trust M = 5.60, SD = 0.43). Participants with only a child of survivor
mother (intimacy M = 6.05, SD = 0.86 and trust M = 6.44, SD = 1.36) and those whose
parents were both children of survivors (intimacy M = 6.26, SD = 0.58 and trust M = 6.61,
SD = 0.54) scored more similarly. None of these differences reached significance though
the results pertaining to intimacy did come close (p = 0.06).
Rubenstein (1981) searched for differences in anxiety between primary school aged
children with one versus two Holocaust survivor grandparents. While none of his results
reached statistical significance it is interesting to see that the effect noted in children of
survivors based on number of survivor parents seems to also be reflected in the third
generation.
Rubenstein’s (1981) group of children with two Holocaust survivor
grandparents (n = 15) recorded higher levels of anxiety, than those with only one (n = 24),
on two measures. The children themselves completed the Fear scale of the Louisville
Behavior Checklist (two survivor grandparents M = 54.20, SD = 13.55; one survivor
© Janine Lurie-Beck 2007
169
grandparent M = 52.14, SD = 9.91), while their school teachers also rated their anxiety on
the School Behaviour Checklist (two survivor grandparents M = 51.92, SD = 10.64; one
survivor grandparent M = 46.72, SD = 9.03).
Sigal and Weinfeld (1989) presented data relating to parental and teacher ratings of
grandchildren of Holocaust survivors on the level of fear/anxiety they evidenced in
ordinary situations. Specifically they reported the proportion of grandchildren reported to
show this symptom often or very often.
They provided data for two groups of
grandchildren of survivors: one group with at least one Holocaust survivor grandparent (n =
58) and one group with at least one survivor parent and one survivor grandparent (n = 11).
Specifically the percentages were:
34% of grandchildren of survivors and 40% of
children/grandchildren of survivors.
9.4. – Summary and Conclusions
This chapter has reviewed and meta-analysed, where possible, the current state of evidence
in relation to the impact of demographic variables on the psychological health of children
and grandchildren of survivors. For the majority of these variables, assessment has been
too sparse or too tainted by methodological problems to make any firm statements about
their impact on descendants of survivors.
While variables such as gender have been studied a number of times, the impact of
variations in the survivor ancestors’ Holocaust experiences have been under-examined to
date. Since we know that such differences led to differing levels of adjustment among
survivors, it follows that the children of the most affected survivors would be the most
affected themselves. However, this is an assumption rather than a statement based on a
large body of hard evidence. The assessment of demographic variables and their impact on
the psychological health of descendants of Holocaust survivors is to date inadequate and is
in need of supplementation.
The assessment of some demographic variables that have been under-examined in
the literature is possible via meta-analytic techniques such as meta-regressions.
This
technique has been used here to supplement the current knowledge regarding gender
differences as well as the influence of age, time lapse variables and post-war settlement
location. However, it is necessary to increase the database of raw data assessments of these
variables. Clearly, even though meta-analytic techniques can go some way in addressing
the gaps in knowledge, further data collection is required to clarify the role of a range of
demographic variables on the psychological health of survivors and their descendants.
© Janine Lurie-Beck 2007
170
Chapter Ten – Refinement of the Model of the Differential Impact of Holocaust Trauma
across Three Generations based on Meta-Analyses
Chapters Eight and Nine collated and synthesised the current state of evidence in relation to
demographic differences among Holocaust survivors and their descendants. The purpose of
this chapter is to revisit the model developed in Chapters Two to Five, in light of the
literature reviewed and meta-analysed in Chapters Eight and Nine.
As this chapter
highlights, the literature in relation to the impact of the Holocaust is lacking in a number of
areas. The final sections of this chapter are dedicated to outlining the hypotheses that arise
from the meta-analytic review of the literature and require further empirical study to
answer.
10.1. – Adequacy of the Assessment of Demographic Differences among Holocaust
Survivors and Descendants in the Literature
Chapters Four and Five of the current thesis summarised the theoretical and anecdotal
conjecture on the influence of demographic variables on the psychological well-being of
Holocaust survivors and their descendants. Chapters Eight and Nine went on to conduct a
review of the existing data relating to these demographic variables. Meta-analytic synthesis
was conducted where viable. Cross-referencing the results of the review of the empirical
studies of demographic differences among survivors (presented in Chapters Eight and
Nine) with the literature and conjecture regarding demographic differences (summarised in
Chapters Four and Five) has been conducted to determine whether these demographic
variables have been empirically assessed as adequately as would be warranted by the
strength of the debate in the literature.
10.1.1. – Adequacy of Demographic Analysis for Holocaust Survivors
The variables addressed in this chapter included a survivor’s age during the Holocaust, their
gender, their country of origin, cultural differences, the reason for their persecution, the
nature or type of experiences/traumas they endured during the Holocaust, their loss of
family members and their post-war settlement location. Upon examination of the empirical
literature it was found that only five of these eight variables had been assessed and the
majority of these had not been assessed adequately.
Studies were found that had
empirically assessed age, gender, nature/type of experiences, loss of family and post-war
settlement location. The results of these studies were reported and meta-analysed (where
possible) in Chapter Eight. Table 10.1 provides a summary for the current state of the
© Janine Lurie-Beck 2007
171
literature for each of the demographic variables theorised to have an impact on survivors’
psychological health.
As can be seen, overall the literature pertaining to demographic differences among
Holocaust survivors is ambiguous at best. The only variable that any statement can be
made about with any real confidence is gender and that is that female survivors are
generally worse off than male survivors.
Table 10.1.
Summary of the current state of evidence of the impact of survivor demographics on survivor psychological health
Variable
Number of
Indirect
Clarity of results
direct
Assessment
assessments
via Metaof variable
Analysis
Nature of Holocaust
20
Overall seems that camp survivors are worse off than non-camp
Experiences
survivors, a number of contradictory results
Loss of family
4
Ambiguous

Gender
1 to 5
Females are worse off than males

Age during the
13
Overall appears to be a positive relationship between age and
Holocaust
symptoms, a number of contradictory results

Time lapse since the
Ambiguous
Holocaust

Post-war settlement
2
Ambiguous
location
Reason for
persecution
Country of Origin
1
Eastern European survivors more troubled than Western
European survivors but more research needed
Cultural Differences
-
10.1.2. – Adequacy of Demographic Analysis for Children of Holocaust Survivors
Demographic differences among children of survivors can be related to facets of their
survivor parents’ experiences (as in Table 10.2) or demographic variables intrinsic to them
(as in Table 10.3). More data is required in relation to the impact of survivor parent
demographics. In addition, the impact of the delay between the war and the birth of
children of survivors has not been directly assessed by any study at all and was analysed for
the first time by the current author via meta-analysis. Analysis of this variable with raw
study data (as opposed to study effect sizes and average age data) is important to clarify if
this result is a true reflection of the influence of this variable. There are two variables that
statements can be made about with a certain degree of confidence and these are gender (that
daughters of survivors are worse off) and number of survivor parents (that children of two
survivor parents are worse off than those with only one survivor parent).
© Janine Lurie-Beck 2007
172
Table 10.2.
Summary of the current state of evidence of the impact of survivor parent demographics on child of survivor psychological
health
Variable
Number of direct
Indirect
Clarity of results
assessments of
Assessment via
variable
Meta-Analysis
Type of survivor parents’
5
Ambiguous
Holocaust experiences
Parental loss of family
2
Positive relationship between negative
symptoms and survivor parent’s loss of family
Survivor parent’s country of
2
Ambiguous
origin
Age of survivor parent
4
Ambiguous
during Holocaust

Post-war settlement
1
Ambiguous
location
Table 10.3.
Summary of the current state of evidence of impact of child of survivor demographics on child of survivor psychological
health
Variable
Number of direct
Indirect
Clarity of results
assessments of
Assessment via
variable
Meta-Analysis

Number of survivor parents
3 to 5
Children of two survivor parents are worse off
than those with only one survivor parent
Gender of survivor parent
1
Children of survivor fathers worse off than
children of survivor mothers

Delay between end of the
0
Negative relationship between delay and
war and birth of the children
negative symptoms
of survivors

Gender
2 to 17
Females worse off than males
Birth order
3
Middle order children worse off
10.1.3. – Adequacy of Demographic Analysis for Grandchildren of Holocaust Survivors
There has been very little research with grandchildren of survivors in any form. Research
considering demographic differences within the grandchild of survivor population has been
even more scant. Four studies were located in total: one that assessed gender differences
and three that considered the number of survivor grandparents or child of survivor parents.
Clearly a larger bank of data is required for gender and number of Holocaust
impacted ancestors. In addition to this are the numerous demographic variables that have
yet to be considered in the grandchildren of survivor literature. These variables include
details of their survivor grandparents’ Holocaust experiences (nature/type of experiences,
loss of family, reason for persecution, country of origin, age during Holocaust, post-war
settlement etc) and demographic factors related to their child of survivor parents (delay
between the end of the war and their birth, birth order). As would be hypothesised for
children of survivors, it would be expected that grandchildren of the most affected
survivors and/or children of the most affected children of survivors will be the most
© Janine Lurie-Beck 2007
173
affected themselves. There is much scope for exploration of demographic differences
within the grandchild of survivor population.
10.2. – Intergenerational Differences within the Holocaust Population
The results and meta-analyses presented in Chapter Seven provide support for the argument
that Holocaust survivors, and their children and grandchildren, evidence higher
levels/greater severity of psychological symptoms than the general population (as
represented by control/comparison groups). However, the question of whether survivor
descendants have been affected to the same extent as the survivors themselves needs to be
addressed also. In other words does the difference between survivors and descendants and
the general population decrease with each generational separation from the Holocaust?
10.2.1. – Direct Intergenerational Comparisons in the Literature
There have been very few studies that have directly compared survivor and descendant
samples. Seven studies were located that collected data that allowed the authors to directly
compare at least one generation to another.
Puzzlingly, statistical analysis of the
differences between generations of survivors were not always carried out by these
researchers. Fortunately most reported enough data/information that these comparisons
could be conducted by the current author.
Of the seven studies that reported multi-
generational data, five reported survivor and children of survivor data, one reported
children and grandchildren of survivor data and the seventh study included all three
generations.
As can be seen by examining Table 10.4, from the small amount of data available it
does appear that survivors overall have scored higher on depression, anxiety and paranoia
and are less likely to view the world as benevolent than children of survivors. The results
for depression, anxiety and world benevolence obtained statistical significance. These
findings tend to support the hypothesised dissipation of the effect of the Holocaust with
each generation.
© Janine Lurie-Beck 2007
174
Table 10.4.
Summary of meta-analyses of survey studies comparing survivors to children of survivors
Variable
Depression
Anxiety
Paranoia
Assumption that
world is
benevolent
Assumption that
world is
meaningful
Number of
results where
survivors
scored higher
than children
of survivors
2/3
2/2
1/1
0/1
0/1
g
95%
confidence
limits for g
Fail
Safe N
Highest individual
result effect size
(g)
Survivor
sample
size
Child of
survivor
sample size
Q
Homogeneity
test
0.55 *
0.38 *
0.12
- 0.45
*
0.33 to 0.78
0.03 to 0.73
- 0.35 to 0.57
0.10 to 0.79
4
<1
-
0.93
0.41
-
151
60
31
67
164
73
44
67
13.24 *
0.04
-
#
-
67
67
-
Note. A positive effect size ( g ) indicates that survivors scored higher on a variable than children of survivors while a
negative effect size indicates that survivors scored lower on a variable than children of survivors. It was not possible to
calculate an effect size for the world assumption of meaningfulness as there was insufficient data, however it can be
stated that the mean score for the children of survivor group was higher than the survivor group, but not statistically
significantly higher.
* p < 0.05
In addition to the results included in the meta-analyses in Table 10.4, there were two
additional studies that reported data from both survivor and children of survivor groups.
Sagi-Schwartz et al. (2003) quoted data pertaining to the rate of insecure attachment among
survivors and children of survivors while Leon, Butcher, Kleinman, Goldberg and Almagor
(1981) presented data on depression and anxiety.
Sagi-Schwartz et al. (2003) reported the percentage of survivors and children of
survivors in their all female sample that obtained an insecure attachment classification. An
analysis testing the difference in percentage between the female survivors and daughters of
survivors was not conducted by Sagi-Schwartz et al. (2003) and so the current author ran
the test using odds ratio analysis. This odds ratio analysis revealed found that of the 48
female survivors, 77% were classified as insecure compared to 54% of the 48 daughters of
survivors. This difference obtained a statistically significant odds ratio of 2.85, meaning
that the odds of female survivors being classified as insecure were 2.85 higher than the
odds of being classified as insecure among daughters of survivors.
Sagi-Schwartz et al.
(2003) also measured anxiety in their study but unfortunately they did not quote the
descriptive data for the female survivors and daughters of survivor groups to allow an
analysis of the difference between them. They only conducted tests comparing each group
to a control group.
Leon et al. (1981) conducted the only other study located to collect data from both
survivors and children of survivors. It was not possible to include their data in the metaanalysis presented in Table 10.4 because they reported their data in terms of 4 sub-groups,
© Janine Lurie-Beck 2007
175
with a total of 8 sub-groups when both generations are included. There was no information
to indicate which groups were related to each other. However, the current author calculated
a mean score for survivors and children of survivors so that at the very least a perusal of the
difference between the means could be made. Leon et al. (1981) used the depression and
paranoia subscales of the MMPI. On the depression subscale the weighted average mean
for the survivors (n=42) was 62.69 compared to 58.00 for the children of survivors (n =
41). For paranoia the pattern was mirrored (survivors M = 57.07; children of survivors M
= 55.24). No significance test is possible here to determine if either of these differences
reached significance but it is notable that in both cases the survivors scored higher (even if
only slightly) than the children of survivors.
Analysis of differences between children and grandchildren of survivors was
afforded by two studies. The results of these studies are presented in Table 10.5. None of
the effect sizes quoted were statistically significant, however the results in relation to
attachment in particular were based on small sample sizes (n = 11 for grandchildren).
Despite the lack of significance, the effect sizes for attachment are in the direction that
would be predicted if further dissipation of the impact of the Holocaust were hypothesised.
In other words, the children of survivors scored higher on negative attachment dimensions
and lower on positive attachment dimensions than grandchildren, suggesting a more
positive attachment picture for the grandchildren than the children of survivors.
Table 10.5.
Summary of meta-analyses of survey studies comparing children to grandchildren of survivors
Variable
Depression
Positive Attachment
Dimensions
Negative Attachment
Dimensions
Number of results where
children scored higher than
grandchildren
0/1
0/1
g
95% confidence
limits for g
Child of survivor
sample size
Grandchild of survivor
sample size
-0.04
-0.14
- 0.33 to 0.25
- 0.78 to 0.50
91
70
91
11
1/1
0.12
- 0.51 to 0.76
70
11
Note. A positive effect size ( g ) indicates that children of survivors scored higher on a variable than grandchildren while a
negative effect size indicates that children scored lower on a variable than grandchildren.
* p < 0.05
In the only study located that collected data from all three generations, Jurkowitz
(1996) found that her survivor group (n = 91) scored statistically significantly higher than
both her children of survivor group (n = 91, t (180) = 6.25, p < 0.001) and her
grandchildren of survivor group (n = 91, t (180) = 6.20, p < 0.001) on the Center for
Epidemiologic Studies Depressed Mood Scale (survivors M = 21.39, SD = 13.90; children
M = 10.41, SD = 9.20; grandchildren M = 10.78, SD = 8.40). Interestingly here the
© Janine Lurie-Beck 2007
176
children of survivors and grandchildren of survivors scored relatively equivalent scores,
rather than evidencing a linear decline in severity of symptoms from survivors down to the
grandchildren.
10.2.2. – Indirect Intergenerational Comparisons via Meta-analysis
Apart from the studies that directly consider the issue of intergenerational differences, an
indirect assessment can also be made by examining the meta-analytic results presented in
Chapter Seven. These meta-analyses combined study results that compared survivors and
descendants to control or comparison groups designed to represent the general population.
The size of the differences between the survivor groups and the control groups can give an
indication of whether the impact of the Holocaust is felt as keenly by survivor descendants
or whether there is evidence of a dissipation of the impact of the Holocaust. If there is a
dissipation of the effect then we would expect to see smaller effect sizes when comparing
descendant groups to control groups than the effect sizes derived when comparing survivor
groups to control groups. In other words, with each generational separation from the
Holocaust, less disparity from the general population is expected.
Table 10.6.
Survivor, child of survivor and grandchild of survivor groups versus control groups – Intergenerational comparison of
meta-analytic effect sizes
Depression
Anxiety
Paranoia
Assumption that would is
benevolent
Assumption that world is
meaningful
Positive Attachment
Dimensions
Negative Attachment
Dimensions
* p < 0.05
Survivors versus
control
effect sizes
95% confidence
limits for g
Children of
survivor versus
control
effect sizes
0.10 *
0.18 *
0.21 *
0.01
95% confidence
limits for g
0.32 *
0.57 *
0.45 *
- 0.28 *
0.26 to 0.39
0.47 to 0.67
0.34 to 0.56
- 0.16 to - 0.40
- 0.11
- 0.23 to 0.00
-
- 0.26 *
- 0.02 to - 0.51
- 0.09
- 0.22 to 0.04
- 0.43 *
- 0.81 to - 0.04
0.28 *
0.03 to 0.52
0.02
- 0.08 to 0.13
0.46
- 0.22 to 1.13
0.01 to 0.19
0.06 to 0.29
0.04 to 0.39
- 0.33 to 0.34
Grandchildren of
survivors versus
control effect
sizes
0.41 *
0.43 *
- 0.23
-
95% confidence
limits for g
0.07 to 0.76
0.15 to 0.72
- 0.74 to 0.27
-
Perusal of Table 10.6 provides some evidence for a dissipation of the impact of the
Holocaust when comparing children of survivor results to survivor results. Indeed, the fact
that the confidence intervals associated with depression and anxiety results for survivors
and children of survivors do not overlap suggest that the effect sizes for children of
survivors on these two variables are statistically significantly lower than for the survivor
generation. Curiously the results for the grandchildren of survivors suggest there may be a
reversal of this trend with severity of symptoms for grandchildren more disparate to the
general population/comparison groups than children of survivors. However, the meta© Janine Lurie-Beck 2007
177
analytic results for grandchildren of survivors are based on much smaller sample sizes
(ranging from 11 to 109 compared to a range of 67 to 1,087 for children of survivor results
and 130 to 2,000 for survivor results). Therefore, it is not clear whether these results
represent a genuine finding of heightened symptoms among grandchildren of survivors or
whether it is merely a statistical artefact.
10.3. – The Need for Further Investigation
In Section 10.1 a review of the moderating role of demographic variables as determined by
meta-analyses in the current thesis was conducted. It is clear that for some demographics it
is possible to determine their impact, however it is also clear that for others the available
data provides only ambiguity or may not even exist at all. Section 10.2 outlined the very
limited data available on intergenerational differences with the Holocaust survivor
population.
Some clarity is needed on a whole range of levels as to the demographic sub-groups
of each generation that are more effected by the Holocaust and the degree to which the
impact of the Holocaust has been transmitted across the generations. To this end, it is
necessary to conduct an empirical study to collect the data needed to provide this clarity.
The second stage of research to be conducted for the current thesis is therefore an empirical
study to supplement and clarify the results of the meta-analyses. Section 10.4 outlines the
hypotheses and aims of the empirical study.
10.4. – Hypotheses for Empirical Study
There is a lot of data to make sense of in the current thesis and the process of doing so has
led to the formation of a large number of hypotheses. The specific hypotheses for the
empirical study have been divided into sections. These are:
• hypotheses about the impact of influential psychological processes on the symptom
levels among survivors and descendants;
• hypotheses regarding the modes of intergenerational trauma transmission; and
• hypotheses about demographic differences among survivors and descendants.
10.4.1. – Hypotheses Regarding the Relationships between Model Variables
10.4.1.1. – The impact of influential psychological processes.
Following on from the literature discussed in Chapter Two the following model hypotheses
(MH) are made about the role of psychological processes in determining the severity or
strength of negative and positive psychological symptoms/dimensions:
© Janine Lurie-Beck 2007
178
MH1: Negative/dysfunctional coping strategies will be positively related to negative
psychological symptoms and negatively related to positive psychological dimensions,
while positive/functional coping strategies will be negatively related to negative
psychological symptoms and positively related to positive psychological dimensions.
MH2: Strength of belief that the world is benevolent and meaningful will be negatively
related to negative psychological symptoms and positively related to positive
psychological dimensions.
Slightly tangential but still somewhat related to this group of hypotheses is a hypothesis
about the relationship between the positive psychological impact of posttraumatic growth
and the negative psychological impacts included in the model.
Studies have found a
positive relationship between posttraumatic growth and pathological symptoms (Cadell et
al., 2003; Laufer & Solomon, 2006; McGrath & Linley, 2006; Morris et al., 2005).
However, Green et al’s (1985) trauma model (Processing a Traumatic Event: A Working
Model), which has been used as a partial basis for the model of Holocaust trauma being
developed in the current thesis, implied a mutual incompatibility of negative
symptomatology and posttraumatic growth. The body of evidence on this issue has been
growing and is sufficient enough that for this study it is hypothesised that among survivors:
MH3:
Posttraumatic growth aspects will co-exist with negative psychological
symptoms (in other words posttraumatic growth and negative psychological symptoms
will be positively related).
10.4.1.2. –The odes of intergenerational trauma transmission.
The way in which Holocaust trauma is revisited in the descendants of survivors was
discussed in Chapter Three where theorised modes of intergenerational trauma transmission
were discussed. Being guided by the literature review process that went into that chapter,
as well as the overarching theories of attachment and the 3-D Circumplex model of family
systems, it is hypothesised that the following variables will be positively associated with
negative psychological symptoms and negatively associated with positive psychological
symptoms:
MH4: Negative parent-child attachment dimensions such as the degree of coldness and
ambivalence will be positively associated with negative psychological symptoms and
negatively associated with positive psychological symptoms and positive parent-child
attachment dimensions such perceived parental warmth will be negatively associated
© Janine Lurie-Beck 2007
179
with negative psychological symptoms and positively associated with positive
psychological dimensions
MH5: A curvilinear/U-shaped relationship will exist between negative psychological
symptoms and family cohesion (with very low and very high cohesion associated with
higher symptom levels than mid-range scores) and an inverted U-shaped relationship
will exist between positive psychological dimensions and family cohesion
MH6: The degree to which parents are encouraging of their children’s attempts to
establish independence will be negatively associated with negative psychological
symptoms and positively associated with positive psychological dimensions
MH7: General communicativeness within the family unit will be negatively associated
with negative psychological symptoms and positively associated with positive
psychological dimensions
MH8: Negative modes of communicating about the Holocaust, such as guilt-inducing,
indirect and non-verbal will be positively associated with negative psychological
symptoms and negatively associated with positive psychological symptoms and positive
modes of communicating about the Holocaust, such as frequent, willing and open
discussion will be negatively associated with negative psychological symptoms and
positively associated with positive psychological dimensions
As well as hypothesising links between these family interaction variables and psychological
impact variables it is also hypothesised that these variables mediate the relationship
between ancestor and descendant scores on psychological impact variables.
10.4.2. – Hypotheses Regarding the Influence of Demographic Variables
For the majority of demographic variables it was possible to make at least a tentative
hypothesis as to which sub-group would be most affected or whether a continuous variable
such as age would be positively or negatively associated with symptoms (if not based on
meta-analytic results then at least on theoretical conjecture outlined in Chapters Four and
Five, as well as Green, Wilson and Lindy’s (1985) Working Model for the Processing of a
Traumatic Event and Wilson’s (1989) Person-Environment Interaction Theory of
Traumatic Stress Reactions).
However, there are a handful of variables where either
insufficient data or very inconsistent data was found in the literature so that it was not
possible to make even a tentative hypothesis at this stage.
© Janine Lurie-Beck 2007
180
The demographic variable hypotheses (DH) are as follows:
DH1: Female survivors, children of survivors and grandchildren of survivors will score
higher on negative variables and lower on positive variables than male survivors,
children of survivors and grandchildren of survivors.
DH2: Scores on negative variables will increase with the age a survivor was during the
Holocaust (as operationalised by their age in 1945) and scores on positive variables will
decrease with age.
DH3: Survivors of camps will score higher on negative variables and lower on positive
variables than survivors with other experiences such as being in hiding or fighting with
partisan groups.
DH4: Scores on negative variables will increase with the amount of a survivors’ family
losses and scores on positive variables will decrease with the amount of family losses
DH5: Survivors who were persecuted for an “understandable reason” such as resistance
to the Nazi regime or committing a “crime” will score lower on negative variables and
higher on positive variables than survivors persecuted for reasons such as race/ethnicity
or religion
DH6:
Scores on negative variables will increase with the amount of time Nazi
persecution occurred in country of origin and scores on positive variables will decrease.
Or at least scores will vary depending on a survivors’ country of origin.
DH7: Scores on negative variables will increase and scores on positive variables will
decrease with increases in the number of ancestors affected by the Holocaust (for
example the number of survivor parents a child of survivors has, the number of child of
survivor parents and survivor grandparents a grandchild of survivors has).
DH8: Middle birth order children will score higher on negative variables and lower on
positive variables than first born children.
DH9: Child of survivor scores on negative variables will decrease with the delay
between the end of the war (1945) and their birth and scores on positive variables will
increase with delay
In addition to these generation-specific hypotheses about demographic differences, it is
hypothesised that:
DH10: The descendants of the most affected subgroups of a particular generation will
be the most affected themselves (for example if camp survivors are more affected than
© Janine Lurie-Beck 2007
181
hiding survivors than children of camp survivors will have higher symptom levels than
children of hiding survivors)
DH11: There will be evidence of a dissipation of the impact of the Holocaust with each
generational separation from the Holocaust. In other words, children of survivors will
score lower on negative variables and higher on positive variables than survivors, and
grandchildren of survivors will score lower on negative variables and higher on positive
variables than children of survivors.
The following variables are hypothesised to have an effect but a directional
hypothesis cannot be made as yet based on the available empirical literature: post war
settlement location, length of time before resettlement of survivors/time spent in DP camps,
gender of survivor or child of survivor parent if only one, birth of children of survivors
before or after survivor parents’ emigration from Europe, birth order (for grandchildren of
survivors).
The relative importance of the demographic variables in the model is also
something that cannot yet be determined from the current literature. There have been no
studies that have made a concerted effort to prioritise demographic variables as to their
importance in determining the impact of the Holocaust. The majority of studies that have
considered demographic variables have analysed them in isolation, or have only analysed a
very small number of demographic variables at a time. Given that this study will be
collecting data on a whole range of demographic variables, attempts will be made to rank
the demographic variables (and also the influential psychological process and family
interaction/transmission variables) in terms of their importance in predicting scores on
psychological impact variables.
10.4.3. – Hypotheses Relating to Membership of Survivor or Descendant of Survivor
Groups
Quite separate from all of the hypotheses about the relationships discussed in the above
sections is a hypothesis relating to the membership of survivor or descendant of survivor
organisations. There is some conjecture in the literature (see Chapter Two) as to whether
membership of a survivor or descendant organisation is reflective of better or worse
adjustment and whether such groups attract better or worse adjusted people. Following on
from the results of the sub-set meta-analyses presented in Chapters Eight and Nine that
addressed this issue the following sample source hypothesis (SH) can be made:
© Janine Lurie-Beck 2007
182
SH1:
Survivors or descendants who are members of a survivor or descendant
organisation will display higher levels of negative symptoms and lower levels of
positive dimensions/variables than non-members.
10.5. – Summary and Conclusions
This chapter has attempted to bring together the findings of the meta-analyses in Chapters
Eight and Nine as well as provide meta-analytic investigation of intergenerational
differences within the Holocaust survivor population. It is clear that more investigation is
needed to more accurately determine the relationships between various psychological
processes and psychological health as well as clarifying the impact of numerous
demographic variables. The preliminary Model of the Differential Impact of Holocaust
Trauma across Three Generations that was progressively built in Chapters Two through to
Five has not been refined that much as a result of the many meta-analyses presented in
Chapters Seven, Eight and Nine (see Figure 10.1). The demographic variables for which
hypotheses could be made are indicated in Figure 10.1 via a notation of the subgroups
hypothesised to be most affected or via a note of the direction of a hypothesised
relationship. They have also been bolded for ease of reference. For example females of
each generation are hypothesised to have high symptom levels and a negative relationship
is proposed between symptom levels and the length of time between the end of the war and
a child of survivor’s birth.
The process of reviewing and meta-analysing the available data on Holocaust
survivors and their descendants has highlighted the gaps in our knowledge about the
psychological impact of the Holocaust and the factors that determine its severity. An
empirical study designed to fill in at least some of those gaps is what will be presented in
the remaining chapters of the current thesis which collectively make up Section C of the
thesis. This study aims to assess all the variables in the model, thereby clarifying the role
of some demographic variables and providing the first analysis of the role of others. In
addition, the model will be refined by determining a ranking of the strength of influence of
all variables in the model in predicting the psychological health of survivors and
descendants.
© Janine Lurie-Beck 2007
183
DH11 – Dissipation of symptom levels with each generational removal from Holocaust experiences
3rd Generation
(Grand-children of Survivors)
2nd Generation
(Children of Survivors)
1st Generation
(Survivors)
Psychological Impacts
of the Holocaust
Depression
Anxiety
Paranoia
PTSD symptoms
Romantic
Attachment
Dimensions
MH3 - Post-traumatic
Growth
•
•
•
•
•
•
•
•
•
•
•
•
•
Depression
Anxiety
Paranoia
Romantic
Attachment
Dimensions
Depression
Anxiety
Paranoia
Romantic
Attachment
Dimensions
Influential
Psychological
Processes
Modes of Intergenerational
Transmission
of Trauma
Demographic Moderators
Holocaust
Survivor Generation
Children of Survivor
Generation
Grandchildren of Survivor
Generation
DH2 - Age during the Holocaust (Positive)
• Time lapse since the Holocaust
DH1 - Gender (Females)
DH3 - Type/nature of Holocaust experiences
(Camps)
DH5 - Reason for persecution
DH4 - Loss of family (Positive)
DH6 - Country of origin
• Post-war settlement location
• Length of time before resettlement/time
spent in displaced persons camps
MH2 - World
Assumptions
MH1 - Coping
Strategies
MH2 - World
Assumptions
MH1 - Coping
Strategies
MH4 - Parent-Child Attachment
MH5 - Family Cohesion
MH6 - Encouragement of
Independence
MH7 - General Family
Communication
MH8 - Communication about
Holocaust experiences
MH2 - World
Assumptions
MH1 - Coping
Strategies
MH4 - Parent-Child
Attachment
MH5 - Family Cohesion
MH6 - Encouragement of
Independence
MH7 - General Family
Communication
DH10 • Age during the Holocaust
• Time lapse since the Holocaust
• Gender
• Type/nature of Holocaust experiences
• Reason for persecution
• Loss of family
• Country of origin
• Post-war settlement location
• Length of time before resettlement/time
spent in displaced persons camps
DH7 - Number of survivor
parents (Positive)
DH9 - Delay between the end
of the war and their birth
(Negative)
• Birth before or after survivor
parent/s emigration
DH8 - Birth order
DH1 - Gender (Females)
DH10 • Age during the Holocaust
• Time lapse since the Holocaust
• Gender
• Type/nature of Holocaust experiences
• Reason for persecution
• Loss of family
• Country of origin
• Post-war settlement location
• Length of time before resettlement/time
spent in displaced persons camps
DH10 • Number of survivor parents
• Delay between the end of
the war and their birth
• Birth before or after survivor
parent/s emigration
• Birth order
• Gender
DH7 - Number of child of
survivor parents
(Positive)
DH8 - Birth order
DH1 - Gender (Females)
Figure 10.1. Empirical Study Hypotheses Marked on the Test Version Model of the Differential Impact of Holocaust Trauma on Three Generations
© Janine Lurie-Beck 2007
184
Section C
Empirical Assessment of the Model of the Differential Impact of Holocaust Trauma across
Three Generations
© Janine Lurie-Beck 2007
185
Chapter Eleven – Empirical Study Rationale and Methodology
The body of the thesis to this point has presented the results of meta-analyses of the
research into Holocaust survivors and their descendants and has presented a model of the
intergenerational transmission of the impact of the Holocaust based on these analyses
(Figure 10.1). However findings relating to a large proportion of this model are inadequate
or ambiguous. It has therefore been necessary to conduct an empirical assessment of this
model to try to determine how accurately it reflects the true relationships between variables
and to reduce some of the ambiguities.
It was felt that the meta-analytic review of existing literature provided sufficient
evidence for the higher rates of pathological symptoms among the survivor and descendant
population compared to the general population. What was clear from a review of this body
of research was that there are definitely differences within the survivor and descendant
population. This was further supported by the meta-analytic findings presented in the
current thesis. Further investigation of differences between survivor and descendant groups
and controls will not really provide any further information or insight. Therefore the focus
of the empirical study reported in the current thesis is on within group differences within
the survivor and descendant populations.
This chapter outlines the rationale for the empirical study conducted as well as the
methodology utilised. Issues relating to sample recruitment and derivation are discussed.
Operationalisation of the variables in the model, including details of reliability and validity
of chosen scales are presented next. Finally design issues, such as the ordering of scales
within the questionnaire booklet, and the statistical analysis approach are explained.
11.1. Rationale of the Empirical Study
Having conducted the extensive review and meta-analyses of the literature regarding
Holocaust survivors and their descendants, it was clear that many potential demographic
and situational moderators had been inadequately assessed. The proposed relationships
between some of the variables in the model, while often discussed in theoretical or
anecdotal literature, were also lacking in empirical verification.
An empirical study that aimed to assess all of the variables reviewed in the metaanalyses would go a long way to clarifying and/or illuminating the influence of some of
these variables.
© Janine Lurie-Beck 2007
186
The aims of the empirical study were to:
• Clarify the nature and strength of relationships between the various theorised impacts of
the Holocaust, namely psychopathological and inter-personal processes.
• Clarify the nature and strength of the moderating influence of demographic and
situational variables on the theorised impacts of the Holocaust.
• Attempt to determine which variables are the most predictive or prescriptive of the most
vulnerable sub-groups of survivors, children and grandchildren.
11.2. – Method
The methodological approach of the empirical study conducted for this thesis is outlined in
detail in this section. Specific information about the design, sample, procedures, measures
and statistical approaches are presented in turn.
11.2.1. – Design
The dependent variables (DVs) in this study are labelled as psychological impact variables
in the model of differential impact of Holocaust trauma across three generations.
Specifically the DVs are depression, anxiety, paranoia/vulnerability, adult attachment
dimensions, post-traumatic stress symptoms and post-traumatic growth dimensions. The
independent variables (IVs) in this study are the influential psychological process variables
of coping strategies and world assumptions, the family interaction process/mode of
transmission variables of parent-child attachment, parental fostering of autonomy, family
cohesion, communication about Holocaust experiences and general family communication.
The demographic variables listed on the model also serve as IVs.
11.2.2. – Sample
The total sample derived for the study was 124 participants from the three generations: 27
Holocaust survivors, 69 children of survivors and 28 grandchildren of survivors.
Ultimately, 32% (40) of the sample was obtained via contacts with survivor or descendant
groups, 10% (12) as a result of media coverage, 6% (7) via Jewish groups, supplemented by
family members (34%, 42) and friends (12%, 15) of participants. There were a small
number of returned questionnaires from an unknown source (6%, 8).
© Janine Lurie-Beck 2007
187
The same criteria used for the meta-analyses were used to define the three generations in
the empirical study:
• A Holocaust survivor was defined as any person who suffered some form of persecution
by the German Nazi Regime/Third Reich. This was dated from January 1933 (when
Hitler came to power) until the end of World War II.
• A child of a Holocaust survivor had to have at least one parent who met the Holocaust
survivor criteria, and had to be born after the cessation of hostilities. Children born
before the end of the war (1945) were only included if their survivor parent/s had
somehow escaped persecution before the end of the war and the child was born after
their escape.
• A grandchild of a Holocaust survivor had to have at least one grandparent meeting the
criteria to be classified as a Holocaust survivor and a parent meeting the criteria for a
child of a Holocaust survivor (and no parent meeting the survivor criteria).
The sources from which the study participants were derived mean that the current
study sample can be classified as a mixed, non-clinical community sample (in terms of the
sample categories referred to in the meta-analyses conducted for this thesis and reported in
Section B). The sample is a mixed community sample in that it contains participants who
are members of survivor or descendant organisations and also participants who are not
members of such an organisation.
Variable scores of members and non-members of
survivor or descendant organisations are compared statistically to determine statistically
significant differences related to group membership. No participants were obtained via
“clinical
sources,”
such
as
client/patient
psychiatric/psychological facilities/practices.
lists
from
in-patient
or
out-patient
While the sample is a convenience rather
than a random sample, the difficulty and cost associated in obtaining such a sample are
beyond the scope of a PhD research study.
11.2.3. – Procedure
At the outset ethical clearance for the empirical study was sought and granted from the
Queensland University of Technology Human Research Ethics Committee (reference
3779H). This process ensured the proper conduct of the study including approval of
methods used to maintain participant anonymity and confidentiality as well as ensuring
participants’ well-being during and as a consequence of their participation.
© Janine Lurie-Beck 2007
188
A variety of recruitment methods were used to obtain participants including
canvassing various groups and organisations for help, word of mouth, personal contacts and
indirect advertising via media coverage. In addition, many participants informed family
members about the study who then also indicated their willingness to participate.
The internet was searched for organisations and agencies which might be able to
help in contacting potential participants. The search terms “Holocaust” and “Shoah” were
entered into a worldwide Google search. In addition the links pages of organisations
identified via the Google search were also perused. The organisations contacted fell into
the four broad categories of:
• Holocaust survivor and/or descendant organisations (both Jewish and non-Jewish)
• Jewish organisations.
• Cultural groups set up for immigrants from countries from which survivors could have
originated (for example Poland).
• Political groups, including communist parties, that may have members who were
persecuted for political reasons.
These organisations were contacted by an email in which the author introduced
herself and the study and asked if they could help either by allowing her to put a notice on
their website and/or in their newsletter and/or send an email or letter to their membership
list.
Appendix H details the organisations that agreed to help and the form of help
provided.
The author received some media attention in January 2005 to coincide with the 60th
anniversary of the liberation of Auschwitz-Birkenau. More coverage was received in May
2005 to coincide with the anniversary of VE Day. Many potential participants contacted the
author as a response to these print and radio interviews. Details of these can be found in
Appendix H.
Potential participants contacted the researcher stating their interest after hearing
about the study. They were then either emailed or posted three questionnaire booklets
(corresponding to the three generations of interest namely survivors, children of survivors
and grandchildren of survivors). In addition, an informed consent information letter was
included (see Appendix I). This included information to help potential participants decide
if they met the criteria for the study and which booklet to complete. By providing a copy of
© Janine Lurie-Beck 2007
189
the questionnaire booklet for each generation, participants had the ability to pass on copies
of the appropriate booklets to family members or acquaintances.
After the initial provision of the questionnaire booklet and informed consent letter, a
monthly reminder was sent, with a maximum of four reminders. Participants were quite
hard to obtain and the sample obtained is the end result of over a year of data collection and
canvassing for participants. The response rate (as a proportion of the total number of
people who were made aware of the study) was quite low.
The response rate as a
proportion of questionnaires forwarded by the author was 77%.
11.2.4. – Translations
Because an international sample was sought for this study it was recognised that numerous
translations may be needed. The information sent out to participants in the English version
of the questionnaire booklet and study information intimated that translations could be
sought, if they knew of anyone who would be willing to participate if a translation was
available.
This was considered particularly relevant in attracting survivors who had
settled/remained in Europe to participate. To this end translations of measures used for the
study that were already in existence were gathered so that a complete translation of the
questionnaire booklet could be obtained at minimal cost (for example the DASS in Dutch,
French and Hungarian, and the IES-R in German).
Because of budget restraints,
translations would only be commissioned if sufficient interest was shown to warrant the
expense (approximately $3,000 (AUD) per language). In order to gauge the need for
translations an introductory paragraph about the study was translated into multiple
languages and sent out with all questionnaire packages (see Appendix J).
Such interest
never eventuated for any European countries despite numerous contacts to organisations
based in Europe.
Ultimately the only full translation of the questionnaire booklet compiled by the
current author was in Hebrew. It was thought that this would yield the largest number of
additional participants, given that a number of organisations based in Israel had agreed to
help reach potential participants. The FES and the IES – R already existed in Hebrew
translation and copies of these were obtained. For the remaining questionnaires, a Hebrew
translation was commissioned. A back-translation was also obtained in an attempt to
ensure the accuracy of the translation. Back-translation involves employing a translator to
translate the documents back into the original language without sighting the original
versions. The back-translation is then compared to the original to check that each statement
190
© Janine Lurie-Beck 2007
or question on each questionnaire asked the same thing or reflected the same thought as the
original. As recommended by Maneesriwongul and Dixon (2004), translations were made
into the native tongue of the translator (i.e., the translation from English to Hebrew was
conducted by a bi-lingual native Hebrew speaker while the translation from Hebrew into
English was conducted by a bi-lingual native English speaker).
Potential participants were told that this Hebrew translation was available in
addition to the English version, and that additional translations may be able to be sourced –
however, all completed questionnaires received used the English version (including three
from Israel and one from Germany).
Therefore, while only English language
questionnaires were completed for this study, the reader should note that reasonable
attempts were made to obtain participants who could not speak English and to ensure that
study participation was accessible for all survivors and descendants who wished to
participate and minimise the language barrier to participation.
11.2.5. – Measures
This section will outline the questionnaires used in the study. Each questionnaire is briefly
summarised and information pertaining to its reliability and validity is also presented.
Figure 11.1 provides a graphical representation of the measured variables in the model and
their respective measures.
Individual Variables
Psychological Impacts
• Depression – Depression
Anxiety Stress Scales
• Anxiety – Depression
Anxiety Stress Scales
• Paranoia – Post
Traumatic Vulnerability
Scale
• PTSD symptoms –
Impact of Events Scale Revised
• Romantic Attachment
Dimensions – Adult
Attachment Scale
• Post-traumatic Growth –
Posttraumatic Growth
Inventory
Variable Types/Classes
Influential Psychological
Processes
• World Assumptions –
World Assumptions Scale
• Coping Strategies – COPE
Modes of Intergenerational
Transmission of Trauma
• Parent-Child Attachment –
Parental Care-giving Style
Questionnaire
• Family Cohesion – Family
Environment Scale – Cohesion
Subscale
• Encouragement of independence –
Parental Attachment Questionnaire
– Parental Fostering of Autonomy
Subscale
• General Family Communication –
Family Environment Scale –
Expressiveness Subscale
• Communication about Holocaust
experiences – Holocaust
Communication Questionnaire
Figure 11.1. Identification of Measures of Variables from Model of the Differential Impact of Holocaust Trauma across
Three Generations used in the empirical study
© Janine Lurie-Beck 2007
191
11.2.5.1. – Depression Anxiety Stress Scales (DASS)
The Depression Anxiety Stress Scales or DASS (S. H. Lovibond & P. F. Lovibond, 1995)
was chosen as the measure for anxiety and depression in this study. A copy of this scale
can be found in Appendix K. The DASS was chosen because it provides scores for both
anxiety and depression, including readily calculable information about severity, while not
burdening the participant with an overly long questionnaire to complete. In addition the
DASS already exists in translation in a number of languages that were potentially useful for
the current study (namely Dutch, French and Hungarian). The DASS is a 42 item, 4-point
likert scale (0-3) measure with the three subscales of anxiety, depression and stress each
containing 14 items. Subscale scores are derived by summing scores for each item that
loads on the subscale, meaning that scores on each subscale can range from a minimum of
0 to a maximum of 42. Severity category scores for the anxiety subscale are normal 0-7,
mild 8-9, moderate 10 -14, severe 15-19 and extremely severe 20 and over. The raw score
cut-offs for the severity categories of depression are as follows: normal range 0-9, mild 1013, moderate 14-20, severe 21-27, and extremely severe 28 and over.
The Cronbach’s α for the anxiety subscale was found to be 0.84 (n = 2,914) during
initial assessment of the scales with an undergraduate student sample, while the depression
(0.91) subscale scored even more favourably (S. H. Lovibond & P. F. Lovibond, 1995, p.
27).
Brown, Chorpita, Korotitsch and Barlow (1997) cite equally high and higher
reliabilities for the DASS in clinical samples. Reliability co-efficients obtained for the
current sample were 0.84 for the anxiety scale and 0.95 for the depression scale.
Convergent validity is evident via the high correlations between the DASS anxiety subscale
and the Beck Anxiety Inventory (r = 0.81, n = 717) and the DASS depression subscale and
the Beck Depression Inventory (r = 0.74) (P. F. Lovibond & S. H. Lovibond, 1995).
The Depression subscale covers the depressive symptom clusters of dysphoria,
hopelessness, devaluation of life, self-deprecation, lack of interest/involvement, anhedonia
and inertia. The Anxiety subscale covers autonomic arousal, skeletal muscular effects,
situational anxiety and subjective experience of anxious affect (P. F. Lovibond & S. H.
Lovibond, 1995).
11.2.5.2. – Impact of Events Scale – Revised (IES-R)
The Impact of Events Scale – Revised or IES-R (D. S. Weiss & Marmar, 1997) was used to
assess the three classes of PTSD symptoms (namely, intrusion, avoidance, and hyper© Janine Lurie-Beck 2007
192
arousal), as it is a well-established and recognised measure in the trauma field (see
Appendix L to view this scale). The IES-R scale is scored on a 5-point likert scale (0-4)
and has 22 items in total. The intrusion and avoidance subscales have 7 items and the
hyperarousal subscale has 6 items. Subscale scores are derived by calculating the mean
score for items that load on the subscale. Subscale scores can therefore range from a
minimum score of 0 to a maximum score of 4. A total score is derived by summing the
three subscale scores and can range from 0 to 12.
The 5-point likert scale asks people to consider how distressing a number of
“difficulties” had been for them from 0 “not at all” to 4 “extremely”. Scores on the IES-R
are interpreted in relation to their position on this likert scale. Severity levels or cut-off
scores have not been formulated for the IES-R and in fact the authors of the scale actively
argue against their use (D. S. Weiss, 2005).
The IES-R is always completed with reference to a traumatic event or series of
events. Weiss (1996) obtained reliabilities for the three sub-scales ranging between 0.77
(hyperarousal) and 0.85 (intrusion and avoidance) with a sample of emergency services
workers, whose reference point was a traumatic incident they had attended. Schreiber et al.
(2004) reported alphas ranging between 0.80 and 0.86 for the three subscales, with a
sample of Holocaust survivors measured before and after open-heart surgery with the
surgery being the reference point. Participants of the current study were asked to think
about whether they had experienced any of the symptoms listed on the scale in relation to
their Holocaust experiences in the past week. Reliabilities obtained with the current sample
were 0.88 for intrusion, 0.86 for avoidance, 0.87 for hyperarousal and 0.95 for the total
IES-R score. In support of its validity, Briere (1997, p. 131) reports that the IES-R reliably
differentiates between traumatised and non-traumatised study participants and reflects the
symptoms of PTSD as listed in the DSM-IV-TR.
11.2.5.3. – Post-Traumatic Vulnerability Scale (PTV)
The Post-Traumatic Vulnerability (PTV) Scale (Shillace, 1994) was used to measure
“paranoia” as referred to in the Holocaust survivor literature. The PTV is a 24 item, truefalse measure. Scores on the PTV can range between 0 and 24, with a score of 1 being
given to any item that is scored in the direction that suggests post-traumatic vulnerability.
The descriptions of paranoia suffered by survivors and descendants in the clinical literature
refer to fear and anxiety about safety and vulnerability of self, friends and family. Paranoia
scales tend to address non-relevant tangential issues such as the perception of mind control,
© Janine Lurie-Beck 2007
193
evil spirits and common phobic reactions. Upon examination, it was felt that the items in
the PTV Scale more accurately reflect the clinical descriptions of paranoia in the survivor
and descendant populations. According to the author, the PTV scale measures:
“A
perceived sense of defencelessness; a sense of insecurity and expectation that danger exists
and harm will occur; overvigilance and caution to protect self and loved ones (Shillace,
1996).” This scale has been found to have relatively high reliability ( α = 0.79) with a
sample of undergraduate students (Shillace, 1996). The PTV scale was found to have a
reliability of 0.51 with the current study sample.
While this reliability is quite low, it
should be noted that the reliability of the scale according to previous research was quite
adequate. Further, according to Streiner (2003), reliabilities between 0.50 and 0.60 are
acceptable for exploratory research. In addition, Shillace (1994) argued for the PTV's
particular usefulness in determining a perceived vulnerability as a result of a traumatic
experience in which defencelessness exists – a characteristic that certainly can be applied to
the experiences of many Holocaust survivors. A copy of this scale can be found in
Appendix M.
11.2.5.4. – Adult Attachment Scale (AAS)
The Adult Attachment Scale developed by Collins and Read (1990), based on Hazan and
Shaver’s (1987) attachment style measure, was used to measure adult/romantic attachment
dimensions (see Appendix N). Hazan and Shaver’s (1987) measure asked participants to
chose between three paragraph descriptions of attachment style corresponding to a secure,
avoidant, or anxious/ambivalent style.
The problem with this approach was that
participants could not choose to endorse only part of these descriptions, if they only agreed
with them in part. Collins and Read (1990) divided these three paragraph into individual
statements making it easier for participants to endorse separate aspects of the three
attachment styles in Hazan and Shaver’s (1987) measure. Collins and Read (1990) also
added three new items to the measure. The AAS is an 18 item, 5-point likert scale (1-5)
measure. The three subscales of the AAS are attachment anxiety, comfort with depending
on others and comfort with closeness with others. Each subscale has 6 items that load on it.
Subscale scores are calculated by summing the scores given to items that load on them.
Subscale scores can therefore range between 6 and 30.
In the current study, a composite
score for “positive attachment dimensions” was also created by summing the scores derived
© Janine Lurie-Beck 2007
194
for the comfort with depending on others, and comfort with closeness with others
subscales. Scores for this composite scale can range between 12 and 60.
This scale was chosen because it provides the opportunity to determine a
participant’s attachment style [as a category within the four category (secure, fearful,
dismissing, or pre-occupied) paradigm of Bartholomew and Horowitz (1991)] as well as the
more recently preferred scale scores of attachment anxiety, and comfort with closeness and
dependency (Brennan, Clark, & Shaver, 1998).
The scoring method used to derive a
participant’s attachment style in the four category paradigm was obtained from one of the
authors of the scale (Collins, 2004). The four category model of attachment as espoused by
Bartholomew and Horowitz (1991) is considered preferable to Hazan and Shaver’s (1987)
three category model as it delineates two styles of avoidant attachment behaviour (that is
dismissing and fearful – combined in the three category model as avoidant). The four
categories of attachment style, as defined by Bartholomew and Horowitz (1991) are
presented in Table 11.1.
Table 11.1
Bartholomew and Horowitz’s (1991) definitions of their four categories of adult attachment
Attachment Type
Definition
Secure
A sense of worthiness (lovability) plus an expectation that other people are
generally accepting and responsive. Comfortable with intimacy and autonomy.
Insecure Attachment Types
Preoccupied
A sense of unworthiness (unlovability) combined with a positive evaluation of
others. This combination of characteristics would lead the person to strive for
self-acceptance by gaining the acceptance of valued others. Corresponds to
Hazan and Shaver’s (1987) ambivalent category
Fearful
A sense of unworthiness (unlovability) combined with an expectation that others
will be negatively disposed (untrustworthy and rejecting). By avoiding close
involvement with others, this style enables people to protect themselves against
anticipated rejection by others.
Dismissing
A sense of love-worthiness combined with a negative disposition toward other
people. Such people protect themselves against disappointment by avoiding
close relationships and maintaining a sense of independence and invulnerability.
According to Collins and Read (1990), the three subscales of the AAS have reasonable
reliability levels.
Cronbach’s α for attachment anxiety was 0.72, for comfort with
dependence was 0.75 and for comfort with closeness was 0.69, with a sample of 406
undergraduate students. For the current study sample the reliability co-efficients were 0.77
for attachment anxiety and 0.85 for the composite scale combining comfort with closeness
with and dependence on others.
The three scales have strong convergent validity,
correlating strongly with corresponding scales of other attachment measures. Sperling,
© Janine Lurie-Beck 2007
195
Foelsch and Grace (1996) found strong and predictably directed correlations (ranging
between 0.50 and 0.84) between the subscales of the AAS and the subscales of the
Attachment Style Measure, the Attachment Style Inventory and the Anxious Romantic
Attachment Style questionnaire providing evidence of convergent validity for the AAS.
11.2.5.5. – Post-Traumatic Growth Inventory (PTGI)
Post-traumatic growth was measured with Tedeschi and Calhoun’s (1996) Post-traumatic
Growth Inventory (PTGI). A copy of this inventory is reproduced in Appendix O. This
measure was chosen as it was the one used by the only other study to examine posttraumatic growth among Holocaust survivors (Lev-Wiesel & Amir, 2003). The PTGI is a
21 item, 6-point likert scale (0-5) measure. This questionnaire is made up of five subscales,
all of which have relatively good reliabilities with an undergraduate student sample: new
possibilities ( α = 0.84), relating to others (0.85), personal strength (0.72), spiritual change
(0.85), appreciation of life (0.67). The total PTGI score has a reliability of 0.90. For the
current sample, the reliabilities were new possibilities (0.87), relating to others (0.89),
personal strength (0.89), spiritual change (0.84), appreciation of life (0.82), and total PTGI
score (0.94). Subscale scores are derived by summing all items that load on them and the
total PTGI is derived by summing all the items. Not all subscales have the same number of
items that load on them so therefore the ranges of possible scores for each differs. Relating
to others has 7 items and so scores can range between 0 and 35, new possibilities has 5
items and a score range of 0 to 25, personal strength has 4 items and a range of 0 to 20,
appreciation of life has 3 items and a range of 0 to 15 and spiritual change has 2 items and a
range of 0 to 10. The total score has a range from 0 to 105.
In the relating to others subscale, participants are asked to consider whether they
have experienced an increased sense of closeness with others, willingness to express
emotions, compassion for others, effort in relationships, knowledge that others can be
counted on in times of trouble, sense that people can be wonderful and that it is okay to
need others. The new possibilities subscale covers positive impacts of trauma such as the
development of new interests, a new path in life, the recognition that better things can be
done with life, the recognition of new opportunities as a result of trauma survival and a
readiness to try to change things that need changing. The posttraumatic growth aspect of
personal strength encompasses an increased sense of self-reliance, acceptance of events,
knowledge that difficulties can be handled and the discovery of strong inner-strength.
© Janine Lurie-Beck 2007
196
Spiritual change aspects include a strengthening of religious faith as well as enhanced
focus on life’s’ priorities. Finally, the appreciation of life subscale addresses issues such as
an appreciation of the value of life and of every day or, in other words, a desire to live life
to its fullest.
The scoring of these subscales requires a summation of scores given to all items.
However, because the number of items that load on each scale differs, a comparison of
subscale scores is not possible in this format. In light of this, average scores for each
subscale were also calculated so that the relative rating given to each subscale could also be
deduced.
11.2.5.6. – COPE – Long Version
Coping strategies were assessed with the aid of the COPE – Long Version (Carver et al.,
1989). The COPE is a 60 item 4-point likert scale (1-4) questionnaire comprising 15
subscales. Subscale scores are derived by summing the 4 items that load on it creating a
score range between 4 and 16. The questions were framed so as to determine dispositional
coping strategies as opposed to being situation-specific. Table 11.2 provides definitions of
the 15 coping strategies as well as an indication of whether they are generally considered to
be a positive or negative influence on mental health by the scale authors (Carver et al.,
1989).
Correlation matrices between the 14 subscales of the COPE and the psychological
impact variables were examined. The coping strategies of behavioural disengagement,
mental disengagement, denial, substance use, focus on and venting of emotions and
religious coping were consistently positively related to negative impact variables and
negatively related to positive impact variables. This grouping is largely consistent with that
made by the authors of the COPE scale (Carver et al., 1989).
The nine remaining
subscales/coping strategies assessed by the COPE are considered positive or functional by
the scales’ authors (Carver et al., 1989) and are consistently negatively related to negative
impact variables and consistently positively related to positive impact variables. These
coping strategies are acceptance, active coping, humour, planning, positive reinterpretation
and growth, restraint, suppression of competing activities and the use of emotional and
instrumental social support. The only departures from Carver et al’s (1989) categorisations
are the classification of religious coping as a negative influence (it was not described as
either negative or positive by Carver et al (1989)) and also focus on and venting of
emotions as a negative influence which was considered positive by the COPE authors
© Janine Lurie-Beck 2007
197
(however they only considered it a positive influence if moderate and acknowledged it
could be a negative influence if extreme). In order to simplify the analyses and aid
understanding it was decided that all analyses considering coping strategies would use a
composite score for “maladaptive coping strategies” and a composite score for “adaptive
coping strategies” based on the groupings derived from the examination of correlation
matrices of data from this study.
Scores for the “maladaptive coping strategies”
composite, with 24 items, can range between 24 and 80 while scores for the “adaptive
coping strategies” composite, with 36 items can range between 36 and 144. The two
composite scales were found to have relatively high reliability with a Cronbach’s α for the
maladaptive composite scale of 0.76 and 0.91 for the adaptive composite scale. No specific
information pertaining to the validity of the COPE could be located. A copy of the COPE
can be found in Appendix P while correlation matrices of the individual coping strategy
subscales with the psychological impact variables can be found in Appendix Q.
Table 11.2.
Definitions and categorisations of COPE subscales
Subscale/
Carver, Scheier and Weintraub (1989)
Coping Strategy
Description of subscales
Categorisation
Behavioural
reducing effort to deal with the stressor, even
Negative
disengagement
giving up the attempt to attain goals with which the
stressor is interfering.
Denial
refusal to believe that the stressor exists or of
Negative
trying to act as though the stressor is not real
Mental disengagement
use of mental strategies to avoid dealing with
Negative
stressor such as day-dreaming
Substance Use
turning to substances as a way of avoiding stressor Negative
Acceptance
accepting reality of situation
Positive
Active coping
taking active steps to try to remove or circumvent
Positive
the stressor or to ameliorate its effects.
Focus on and venting of
focus on distress experiencing and ventilating
Positive
emotions
those feelings.
Humour
making light of the stressor by joking about it
Positive
Planning
coming up with action strategies, thinking about
Positive
what steps to take and how best to handle the
problem.
Positive reinterpretation
construing a stressful transaction in positive terms
Positive
and growth
Restraint
waiting until an appropriate opportunity to act
Positive
presents itself, holding oneself back and not acting
prematurely.
Suppression of
putting other projects aside and trying to avoid
Positive
competing activities
becoming distracted by other events
Use of emotional social
seeking moral support, sympathy or understanding Positive
support
Use of instrumental
seeking advice, assistance or information
Positive
social support
Religious coping
tendency to turn to religion in times of stress
Not classified
© Janine Lurie-Beck 2007
Current Study
Categorisation
Maladaptive
Maladaptive
Maladaptive
Maladaptive
Adaptive
Adaptive
Maladaptive
Adaptive
Adaptive
Adaptive
Adaptive
Adaptive
Adaptive
Adaptive
Maladaptive
198
11.2.5.7. – Benevolence and Meaningfulness of the World sub-scales of the World
Assumptions Scale (WAS)
The benevolence and meaningfulness of the world subscales of the World
Assumptions Scale (Janoff-Bulman, 1996) were utilised in this study (see Appendix R). A
belief in the benevolence of the world is the belief that the world (in both its people and its
events) is a kind and caring place (Janoff-Bulman, 1992 p. 6). A person believes the world
is meaningful if they believe there is a predictable and understandable relationship between
a person’s actions and what events befall them. In other words, good things happen to
good people and bad things only happen to bad people or people who have behaved in such
a way as to deserve a negative consequence (Janoff-Bulman, 1992 p. 8). Both subscales
have been quoted as having quite reasonable reliabilities (Janoff-Bulman, 1996
benevolence = 0.87, meaningfulness = 0.76) with an undergraduate student sample. In the
current study the reliability co-efficients derived were 0.83 for benevolence and 0.75 for
meaningfulness. No validity information could be located for this scale. Participants are
invited to indicate their agreement or disagreement with 20 statements on a 6-point likert
scale (1-6). The benevolence subscale has 8 items while the meaningfulness subscale has
12 items. Subscale scores are derived by summing item scores (with a total of 6 items that
need reverse scoring overall).
Benevolence scores range between 8 and 48 while
meaningfulness scores range between 12 and 72.
11.2.5.8. – Parental Care-giving Style Questionnaire (PCS)
The Parental Care-giving Style Questionnaire (PCS) asks participants to consider
their parent’s attachment behaviour towards them as they were growing up and so were
completed from the “child’s” perspective of the parent-child dyad (see Appendix S). The
PCS is a 12 item, 5-point likert scale (0-4) measure. The PCS is based on Hazan and
Shaver’s (1986, unpublished, cited in Collins & Read, 1990) paragraph descriptions of the
three
parental
care-giving
styles
of
warm/responsive,
cold/rejecting
and
ambivalent/inconsistent. The three subscales, based on these paragraphs, each have 4 items
that load on them and since subscale scores are derived by summing, subscale scores range
between 0 and 16. A warm or responsive care-giving style as rated by the PCS is described
as warm, responsive, supportive and comfortable. A cold/rejecting style is characterised as
cold, distant, rejecting, non-responsive and a seeming lack of concern. An ambivalent or
inconsistent care-giving style is characterised by noticeably inconsistent reactions and
warmth, inability to show love, and seeming pre-occupation with own agendas.
© Janine Lurie-Beck 2007
199
The paragraph descriptions of parental care-giving style were separated into single
statements and rated on a 5-point likert scale, akin to the method used by Collins and Read
(1990) in adapting Hazan and Shaver’s (1987) adult attachment descriptions. This likert
scale version of the questionnaire was first used by Walisever (1995). Walisever (1995)
did not cite any reliability data for the PCS she constructed and so the data from the current
sample was analysed to provide information about the measure’s reliability. As can be seen
in Table 11.3 all three subscales (both mother and father version) obtained quite high
reliability co-efficients. No validity information for this scale was reported by Walisever
(1995).
Table 11.3.
Reliability analysis results for the Parental Care-giving Style questionnaire
Subscale
Mother Warm/Responsive
Mother Cold/Rejecting
Mother Ambivalent/Inconsistent
Father Warm/Responsive
Father Cold/Rejecting
Father Ambivalent/Inconsistent
Cronbach’s α
0.92 (n =95)
0.82 (n = 95)
0.87 (n = 94)
0.90 (n = 91)
0.80 (n = 91)
0.82 (n = 91)
11.2.5.9. – Parental Fostering of Autonomy Subscale of the Parental Attachment
Questionnaire (PAQ-PFA)
The degree to which parents encourage their children’s independence/autonomy is often
used as an indicator of the success of the children’s separation-individuation in the
adolescent and young adult years. The measure chosen to address this issue in the current
study was the Parental Fostering of Autonomy Subscale of Kenny’s (1987) Parental
Attachment Questionnaire (PAQ-PFA). The PAQ-PFA was chosen because after perusing
the individual items/statements on numerous scales purporting to assess this issue, the
PAQ-PFA’s items most closely corresponded to descriptions of survivor and descendant
relations in the literature. The PAQ-PFA is a 14 item, 5-point likert scale (1-5). Scores for
the PAQ-PFA are derived by summing the scores given to each of the 14 items (after 7 of
the negatively worded statements have been reverse scored). Scores can range between 14
and 70. Participants were asked to reflect the degree to which their parents encouraged
their independence during adolescence and early adulthood. Kenny (1987) obtained a
Cronbach’s α of 0.88 for this subscale. However, this figure is based on the version of the
scale that considers both parents together. In this study, it was of interest to consider each
parent, mother and father separately. Therefore reliability analysis was conducted with the
© Janine Lurie-Beck 2007
200
current study sample to provide reliability indicators for the scale when parents are
considered separately. The Cronbach’s α ’s obtained from the current study data were 0.93
when referring to a participant’s mother and 0.91 when based on a participant’s father (n =
95). Heiss, Berman and Sperling (1996) found this scale to be strongly negatively related
to scores on the parental over-protectiveness scale of the Parental Bonding Inventory,
attesting to its divergent validity. The Parental Attachment Questionnaire was viewed
positively in a review of measures designed to assess separation-individuation (Lopez &
Gover, 1993). A copy of this measure can be found in Appendix T.
11.2.5.10. – Lichtman Holocaust Communication Questionnaire (HCQ)
Researchers have measured communication about Holocaust experiences in a myriad of
ways. Therefore, the majority of research to date has measured this variable with adhoc,
constructed questionnaires rather than established questionnaires. Rather than construct a
new measure/questions as many have done, the current author chose to use the
questionnaire developed by Lichtman (1983), as it has been subsequently utilised in a
number of studies and appeared to have good construct/face validity in terms of the issues it
addresses. No reliability or validity statistics have ever been quoted for this scale. The
subscales and the reliabilities obtained with the sample of children of Holocaust survivors
in this study are presented in Table 11.4.
Table 11.4.
Reliability co-efficients for the Holocaust Communication Questionnaire and subscales
Subscale
Total scale
Father’s affective communication about the Holocaust, as conveyed by father
Father’s frequent and willing discussion of his wartime experiences and the
transmission of factual information.
Father’s guilt-inducing communication
Father’s indirect communication about the Holocaust, as conveyed by father
Mother’s affective communication about the Holocaust, as conveyed by mother
Mother’s frequent and willing discussion of her wartime experiences and the
transmission of factual information.
Mother’s guilt-inducing communication
Mother’s indirect communication about the Holocaust, as conveyed by mother
Awareness of the Holocaust at a young age and its nonverbal presence in the home,
as conveyed by either parent
Number
of items
19
1
3
Cronbach’s
α
0.80 (n = 44)
Not computable
0.80 (n = 62)
2
1
1
3
0.36 (n = 61)
Not computable
Not computable
0.77 (n = 55)
2
1
3
0.61 (n = 57)
Not computable
0.47 (n = 68)
As can be seen, the reliability of the scales is quite variable with a number obtaining
quite reasonable reliability levels, while two scales have quite low reliabilities. In defence
of its continued use in the study, it should be noted that these reliability issues could not
have been determined prior to data collection. It should also be noted that the scales are
© Janine Lurie-Beck 2007
201
mostly made up of a very small number of items which has been shown to have a negative
influence on Cronbach’s alpha calculations (Hair, Anderson, Tatham, & Black, 1998, p.
118). Certainly any future research study using this scale should consider attempting to
revise this measure to increase its reliability. However, given the hypothesised central role
played by communication about the Holocaust it was not possible to exclude the measure
from analyses in this study. A copy of this questionnaire can be found in Appendix U.
11.2.5.11. – Cohesion and Expressiveness Subscales of the Family Environment
Scale (FES)
The Cohesion subscale of the Family Environment Scale (Moos, 1974) was used to
measure family cohesion. The questions were phrased retrospectively, and participants
were asked to consider their family of origin. This subscale comprised 9 statements to
which a true or false answer was required. A raw score is derived by giving any positive
item as true and any negative item as false.
This raw score is then converted to a
standardised score with the help of tables provided by Moos (1974). Standard scores for
the cohesion subscale range between 1 and 68. Tutty (1995) cites reliability co-efficients
ranging between 0.77 and 0.86 for this subscale. Moos (1990) cites reliability co-efficients
for this subscale ranging from 0.76 to 0.79 among a variety of samples (including
distressed families). With the current sample a reliability co-efficient of 0.40 was obtained.
While it is granted that this reliability is far from ideal, again this could not have been
predicted apriori. Given the key role family cohesion is hypothesised to play within the
family dynamics and their role in trauma transmission its omission from the analysis
would have had strong repercussions. As cited earlier, Streiner (2003) was of the view that
lower reliabilities are acceptable when the research is exploratory in nature. In further
support of this scales use, strong positive correlations between scores on the FES cohesion
scale and the affective involvement scale of the Family Assessment Device (r = 0.68, p <
0.001) , the enmeshment/disengagement scale of the Structural Family Interaction Scale –
Revised (r = 0.89, p < 0.001 ) the cohesion scale of the Family Adaptability and Cohesion
Evaluation Scale III (r = 0.86, p < 0.001) as found by Perosa and Perosa (1990) more than
attest to this scale’s convergent validity.
Another subscale of the Family Environment Scale (Moos, 1974), Expressiveness,
was used to measure general communication within the family unit. As with the Cohesion
subscale, the Expressiveness subscale comprised 9 true/false statements.
The
expressiveness subscale is scored in the same way as the cohesion subscale with the
© Janine Lurie-Beck 2007
202
exception that standard scores range between 15 and 73. Again, participants completed this
scale in relation to their family of origin. Tutty (1995) suggests the reliability of this scale
ranges between 0.46 and 0.63. Moos (1990) quotes co-efficients between 0.58 and 0.69
from various samples (including various types of distressed families) for this subscale.
With the current sample, the reliability co-efficient was 0.84. No information regarding
validity of this measure was located.
The FES was chosen as a measure for family communicativeness and cohesion
because it has been used in a number of studies with the Holocaust survivor population. By
using only the two subscales of interest, there was the added benefit of minimising the
number of items that participants had to complete within the questionnaire booklet. A copy
of the two FES subscales used could not be appended due to copyright restraints.
11.2.5.12. – Control questionnaire for historical influences.
The data collection stage for this study began shortly before the Indian Ocean Tsunami
which occurred on 26 December 2004.
Given that such a large scale event had the
potential to alter many people’s mood when completing the questionnaires, a control
questionnaire was devised. This questionnaire asked participants to list any historical
events that they felt had a negative impact on them at the time they were completing the
questionnaires. They were also asked to give a rating between 1 and 10 of the strength of
that impact.
On the same questionnaire, participants were asked to nominate any
significant personal events in a similar manner. A copy of this questionnaire can be found
in Appendix V. No participant indicated undue negative affects by contemporary historical
events that would preclude their inclusion in the study sample.
11.2.5.13. – Omission of unresolved mourning measure.
There are very few measures of unresolved mourning in existence. The one study
located that used such a measure (specifically the Grief Measurement Scale) with a
survivor sample concluded that there were problems with its construct validity, particularly
as it applied to the specific Holocaust survivor population (Chayes, 1987). In addition, it
was also thought that survivor respondents may have a response bias when completing such
a measure to reflect perhaps a high degree of unresolved mourning so as to not dishonour
the memory of their family members who died during the Holocaust. Such measures ask
the respondent to contemplate issues such as how often they think about the lost loved one
which are questions which would obviously elicit a biased response, if honouring the
© Janine Lurie-Beck 2007
203
memory of loved ones was of concern. Therefore the issue of unresolved mourning was
not directly measured in the study but, the variables theorised to be affected by it, are.
11.2.5.14. – Ordering of questionnaires in questionnaire booklets.
Table 11.5 outlines which questionnaires were completed by each generation and the
ordering of those questionnaires within the questionnaire booklets.
Each participant
completed a demographic questionnaire about themselves, and children and grandchildren
were asked to fill out demographic questionnaires for any ancestors applicable to the study
who were not participating themselves. In this way the influence of ancestral demographic
variables could be examined for all participants even if their parents or grandparents did not
participate in the study. Copies of the demographic questionnaires appear in Appendix W.
Within the three generation questionnaire booklets, questionnaires were ordered so that the
more potentially distressing scales such as the DASS and the IES-R were separated. While
the ordering of questionnaires was largely the same for the children and grandchildren of
survivors, the ordering was different for survivors. This was because there were more
questionnaires about symptoms which needed to be spread out amongst fewer more
positive questionnaires.
Table 11.5.
Order and content of questionnaire booklets
Holocaust Survivors
Children of Holocaust Survivors
Adult Attachment Scale
COPE
Depression Anxiety Stress Scales
World Assumptions Scale –
Meaningfulness and Benevolence of
the World Subscales
Shillace PTV Scale
Impact of Events Scale - Revised
Post-traumatic Growth Inventory
Participant Demographics
Open Ended Page
Parental Care-giving Style
Questionnaire
Parental Attachment Questionnaire –
Parental Fostering of Autonomy
Subscale
Family Environment Scale –
Cohesion and Expressiveness
Subscales
Lichtman’s Holocaust
Communication Questionnaire
Depression Anxiety Stress Scales
COPE
Adult Attachment Scale
Shillace PTV Scale
World Assumptions Scale –
Meaningfulness and Benevolence of
the World Subscales
Participant Demographics
Open Ended Page
Survivor Parent Demographics
© Janine Lurie-Beck 2007
Grandchildren of Holocaust
Survivors
Parental Care-giving Style
Questionnaire
Parental Attachment Questionnaire –
Parental Fostering of Autonomy
Subscale
Family Environment Scale –
Cohesion and Expressiveness
Subscales
Depression Anxiety Stress Scales
COPE
Adult Attachment Scale
Shillace PTV Scale
World Assumptions Scale –
Meaningfulness and Benevolence of
the World Subscales
Participant Demographics
Open Ended Page
Child of Survivor Parent
Demographics
Survivor Grandparent Demographics
204
11.3. – Description of Sample Obtained
11.3.1. – Description of Holocaust Survivor Sample
A total of 27 Holocaust survivors participated in the study of whom 14 (52%) are male and
13 (48%) are female. Their average current age is 76.93 years with a range of 62 to 100
years. Their average age during the Holocaust (as operationalised by their age in 1945) was
17.33 years with a range of 2 to 41 years.
Education levels of the participants are as follows: 4% (1) have attained
primary/elementary school level, 52% (14) have attained high school level and 44% (12)
have some form of tertiary qualifications. Close to nine out of ten participants (24, 89%)
identify themselves as Jewish, with 7% (2) reporting they are atheist and 4% (1) Christian.
In terms of marital status, 52% (14) are married, 37% (10) are widowed and 11% (3) are
divorced or separated.
In terms of country of residence, 56% (15) currently live in Australia, 30% (8) in
America, 7% (2) in England, 4% (1) in Germany and 4% (1) in Israel. The average amount
of time that survivors spent in Europe after the war before emigrating was 5.17 years.
There is a relatively good range when perusing the country of birth data with 26%
(7) having been born in Austria, 19% (5) in Hungary, 15% (4) in Poland, 11% (3) in the
Netherlands, 11% (3) in Germany, 7% (2) in Belgium, 7% (2) in Lithuania, and 4% (1) in
Latvia.
With regard to their experiences during the Holocaust, 44% (12) spent time in either
a concentration or labour camp. A number indicate that they had been in hiding either as a
child (22%, 6) or as an adult (7%, 2) with the help of false papers or under an assumed
identity. One survivor participant indicated they had been part of the resistance or a
partisan group.
No participant indicated that they had endured any medical
experimentation during the Holocaust. In the immediate aftermath of the war, 37% (10)
spent some time in a displaced persons camp.
It should be noted that 22% (6) of the survivor sample managed to escape Nazi
persecution before 1945 and so did not endure the full extent of traumas suffered by others
from their country of origin. Three of these participants are from Germany with one
escaping in 1937 and two in 1939. The other three participants who escaped are from
Austria with two escaping in 1939 and one in 1941. Given that Nazi persecution of Jews in
Germany started in 1933 when Hitler came to power and in 1938 after the Anschluss in
Austria, it is valid to state that all of these six participants endured some persecution.
205
© Janine Lurie-Beck 2007
Therefore, while they did escape some of the more severe forms of traumatisation, they did
endure the beginning stages of what has become known as the Holocaust and so can still be
categorised as Holocaust survivors. Differences between this group of 6 survivors (who
escaped prior to 1945) and other survivors in this sample are assessed as part of the
analyses presented in this chapter.
While 19% (5) survivor participants believe they are the sole survivor of their
family, 44% (12) indicate that they were alone (without family members) during at least
part of the Holocaust. In terms of specific losses of family members, 44% (11) say their
mother was killed, 39% (10) say their father was killed, with 27% indicating both parents
had been killed. Almost six in ten of those who had siblings (12, 44%) indicated that at
least one of their siblings had been killed (7, 58%). Of the six participants who had a
spouse, two (33%) had been killed. Of the participants who still had living grandparents at
the outset of the war (14, 52%), 79% (11) indicate that at least one of their grandparents
had been killed during the war.
The reason for persecution was included as potentially influential variable in the
Model of the Differential Impact of the Holocaust; however the sample of survivors
obtained are all Jewish (although the reader should note the current author’s attempts to
obtain non-Jewish survivors in Appendix H). Therefore analysis of Jewish versus nonJewish survivors is not possible. Almost eight in ten (21, 78%) indicate that they still
actively participate and practise their Jewish faith.
Just over four in ten (11, 42%) of the survivor sample are members of some kind of
survivor organisation. Just over a quarter (7, 27%) report they have been in individual
therapy at some stage in their life compared to 73% (19) who did not report any therapy
participation. No survivor participants indicate participation in any form of group therapy.
11.3.2. – Description of Children of Holocaust Survivor Sample
There are 69 children of survivors in the sample obtained for this study, with 24 males
(35%) and 45 females (65%). Their average age is 51.22 years and ranges between 29 and
61. The vast majority have a tertiary education (94%, 65), with 4% (3) attaining a high
school certificate and one person not responding to the question. The most frequently cited
religion is Judaism (87%, 60), followed by none/atheism (9%, 6), Christianity (1%, 1) and
other/not specified (1%, 1). Finally, with regard to marital status, 61% (42) reported that
they are currently married, 17% (12) divorced/separated, 10% (7) single/never married, 7%
(5) defacto, and 3% (2) widowed.
© Janine Lurie-Beck 2007
206
The vast majority of participants in this sample have two survivor parents (51,
74%). Twelve (17%) of the 69 children of survivors have a survivor father only and six
(9%) have a survivor mother only. The average time lapse/delay between the end of the
war (1945) and the birth of children of survivor participants was 8.26 years, with a range of
0 to 31 years. One participant was born while their survivor parent/s were in a displaced
persons camp, however 27 (39%) were born in Europe before their parents immigrated to
other continents.
Almost a third (20, 29%) of the child of survivor sample were born in Australia,
with 21% (14) in America, 12% (8) in Poland, 9% (6) in Hungary, 9% (6) in Israel, 4% (3)
in Czechoslovakia, 4% (3) England, 3% (2) in Austria, 3% (2) in the Netherlands, 2% (1)
in Canada, 2% (1) in Denmark, 2% (1) in France, and 2% (1) in Germany. One participant
did not indicate their country of birth. Over half (42, 61%) of the children of survivor
sample currently live in Australia, followed by 28% (19) in America, 6% (4) in New
Zealand, 3% (2) in Canada, and 1% (1) in Israel.
Among the child of survivor sample, 32% have never been in any form of
therapy/counselling, 65% had been in individual therapy, 26% have been in group or
relationship therapy with 24% indicating they have been in both individual and group
therapy. Almost half (45%) report being a member of some form of survivor descendants’
organisation.
The average age of survivor mothers in 1945 of the child of survivor sample was
24.18 and ranged between 5 and 41 years. The average age of survivor fathers in 1945 of
the child of survivor sample was 27.84 and ranged between 2 and 49 years. There was a
reasonable spread of survivor parent Holocaust experiences: 69% of mothers and 70% of
fathers were camp survivors, 22% of mothers and 23% of father were in hiding or had some
other non-camp Holocaust experiences and 10% of mothers and 7% of fathers endured part
of the Holocaust but had escaped before 1945 to another country.
11.3.3. – Description of Grandchildren of Holocaust Survivor Sample
Among the 28 grandchildren of survivors who participated in the study, 8 (29%) are male
while 20 (71%) are female. The average age of grandchildren participants is 25.36 years,
with a range of 20 to 34 years. For 36% (10), both parents are children of survivors, 46%
(13) have a child of survivor mother only and 18% (5) have a child of survivor father only.
In terms of number survivor grandparents, 18% (5) have one, 46% (13) have two, 4% (1)
has three and 32% (9) have all four grandparents as survivors.
© Janine Lurie-Beck 2007
207
Two thirds (18, 64%) of the grandchildren sample were born in Australia, followed
by 14% (4) in America, 11% (3) in Israel, 4% (1) in the Netherlands, 4% (1) in New
Zealand and 4% (1) in South Africa. Almost three quarters (20, 71%) of the grandchildren
in the study currently live in Australia, with a further 21% (6) in America, 4% (1) in Israel
and 4% (1) in New Zealand.
Less than six out of ten grandchildren of survivors in the sample have been in any
form of therapy: 42% (11) have been in individual therapy and 15% (4) have been in both
group and individual therapy, leaving 42% (11) with no therapy experience. Just over one
in ten (12%, 3) indicate they are a member of some form of descendant of survivors
organisation.
As far as education level goes, 4% (1) say they have no formal education, 11% (3)
have a high school education while the vast majority have obtained some form of tertiary
education (86%, 24). Not surprisingly, given the age profile of this group, 89% (25) report
their marital status as single/never married with the remainder living in some form of
relationship (defacto 7% (2) and married 4% (1)). One quarter (75%, 21) identify as Jewish
with 14% (4) reporting their religion as none or atheist and 11% (3) citing some other nonspecified religion.
11.4. – Statistical Analysis Approach
Because a large focus of the thesis and the study is to look at demographic subgroups
within the survivor and descendant population, it is necessary to stratify the sample in many
ways. In so doing, the sample sizes in these sub-samples are often very small (n < 10).
Having to work with small sub-samples has several repercussions for the statistical analyses
that can be validly conducted.
Firstly, it should be noted that while it would be desirable to test the multiple
relationships presented in the model via multivariate analyses this was not at all possible.
The sample size requirements for such tests were prohibitive.
Therefore, a largely
univariate or bivariate approach has been used out of necessity. The statistical tests that are
used include t tests, ANOVAs, χ 2 tests and zero-order bivariate correlations. Statistical
tests designed to measure the relationship between two variables with a third potentially
influential or confounding variable controlled for (namely ANCOVA, mixed design
ANOVA and partial correlations) were also used so that inter-relating IVs were taken into
© Janine Lurie-Beck 2007
208
account as much as the sample sizes and available statistical tools (that could be used
validly with the small samples) would allow.
The determination of the relative importance of model variables in predicting
survivor and descendant scores on psychological impact variables was one of the key aims
of the empirical study. Ideally a multivariate approach would be used to ascertain the
relative importance/rank-ordering of variables. However, because it was not possible to do
this validly (because of the small sample sizes derived) another method of ranking was
used. The proportion of variation in psychological impact variables accounted for by all
remaining model variables (namely influential psychological processes, family interaction
variables and demographic variables) was calculated in the form of η 2 or r 2 . These
η 2 and r 2 were then ranked in order to determine a rank ordering of each class of
variables. Ranking variables by effect sizes in this way as opposed to using significance
level as a ranking tool means that the ranking is not unduly influenced by varying sample
sizes.
Small sample sizes often lead to a decreased ability to obtain statistically significant
results because of a lack of power. When sample sizes are small, much larger differences
between groups, or much larger correlations are required before the significance threshold
is reached. The traditional significance level of 0.05 is set to keep the incidence of Type I
errors (incorrectly rejecting the null hypothesis) to a minimum. When power is low the
incidence of Type II errors (incorrectly retaining the null hypothesis) is increased. While
Type I errors are certainly undesirable, Type II errors are by no means a more palatable
alternative.
There are three main ways to increase power and decrease the likelihood Type II
errors: increase the sample size, increase the size of the effect or increase the significance
level (Keppel & Wickens, 2004; Tilley, 1999). The difficulties in obtaining the sample
derived for the study are evidence that it would be too time-consuming and costly to
increase the sample size. Increasing the effect size is something that largely relates to
experimental manipulations where it is relatively easy to increase the differentiation
between experimental groups by manipulating levels of stimuli, but is not something that is
applicable in this case. This leaves the third option of increasing the significance level as
the only real option left in this case.
© Janine Lurie-Beck 2007
209
To decrease the incidence of Type II errors a more liberal significance level can be
used so that the threshold for a statistically significant result is easier to reach. It was
decided that the threat of Type II errors was quite high for this study, given the large
number of analyses based on small samples, and that the significance level of 0.10 be used
to counter this. Results that have a probability of less than 0.10, but more than 0.05, are
clearly delineated in the text. Results based on samples of 30 or more (or where all groups
in the analysis are more than 10) use only the 0.05 significance level.
The use of the more liberal significance level of 0.10 is also more palatable because
this study is largely exploratory in nature. In many cases, it assesses variables that have not
been assessed before. It is within such a spirit of exploration that a more liberal approach
to statistical significance is perhaps more permissible. The writer is, however, aware of the
limitations of such an approach and recognises and points out to the reader, that larger
samples and further testing are required to validate any findings reported herein.
A small note is also required in relation to the assumptions of the parametric
statistical tests used. The homogeneity of variance assumption, applicable to the t-tests and
ANOVAs reported in the current thesis, was breached numerous times. In the case of ttests, SPSS reports test results that are based on the assumption of equal variances as well
as the assumption of unequal variances. When the homogeneity of variance assumption has
been breached for a t-test result, the results of the t-test which control for the breach (the
version that does not assume equal variances) are reported where the test remains
significant. This will be obvious to the reader via the reporting of adjusted degrees of
freedom. In the case of ANOVAs, the non-parametric equivalent of an ANOVA, namely
the Kruskal-Wallis test, is used.
For the sake of brevity, statistics assessing the already well-established relationships
between the negative psychological symptoms in the psychological impacts column of the
model (i.e., depression, anxiety, paranoia, PTSD symptoms, and insecure attachment) are
presented in Appendix X. Correlation matrices for the influential psychological process
variables as well as the family interaction variables have also been appended (Appendices
Y and Z respectively). These statistics are not required to answer any hypotheses. The
reader should note that these correlations are all in the direction they would be expected to
be.
© Janine Lurie-Beck 2007
210
Chapter Twelve – Empirical Assessment of Influential Psychological Processes and Modes
of Intergenerational Transmission Modes among Survivors and Descendants
In this chapter, the version of the model of the differential impact of Holocaust trauma
across three generations put forward at the end of Chapter Three (before the demographic
variables were added) is assessed empirically. This section of the model is focussed on
establishing the relationships between symptoms among survivors and descendants and the
proposed modes of trauma transmission that attempt to explain why descendants suffer
similarly to the survivors themselves.
Figure 12.1 reproduces the section of the model which will be assessed in this
chapter.
The variables of world assumptions and coping strategies which have been
labelled the influential psychological processes in the model are proposed as factors that
will be correlated with the severity of symptoms/psychological impacts experienced by the
survivors and descendants. The family interaction variables of parent-child attachment,
family cohesion, encouragement of independence, general family communication and
communication about Holocaust experiences are the proposed modes by which Holocaust
trauma is transmitted intergenerationally. In a statistical sense, they are argued to mediate
the relationship between ancestor (for example Holocaust survivor) and descendant (for
example child of survivors) scores on psychological impact variables.
© Janine Lurie-Beck 2007
211
Psychological Impacts
of the Holocaust
Influential
Psychological
Processes
Depression
Anxiety
Paranoia
PTSD symptoms
Romantic Attachment
Dimensions
• Post-traumatic
Growth
• World Assumptions
• Coping Strategies
•
•
•
•
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
• World Assumptions
• Coping Strategies
•
•
•
•
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
• World Assumptions
• Coping Strategies
3rd Generation
(Grand-children of
Survivors)
2nd Generation
(Children of Survivors)
1st Generation
(Survivors)
•
•
•
•
•
Modes of Intergenerational
Transmission
of Trauma
•
•
•
•
•
Parent-Child Attachment
Family Cohesion
Encouragement of Independence
General Family Communication
Communication about Holocaust
experiences
•
•
•
•
Parent-Child Attachment
Family Cohesion
Encouragement of Independence
General Family Communication
Figure 12.1. Section of the Test Model of the Differential Impact of Holocaust Trauma on Three Generations to be tested
in this chapter
In order to test the veracity of the full mechanisms of this model, it will be
necessary to establish statistically the following five relationships:
1. that the influential psychological processes are correlated with the severity or level of
psychological impact variables among survivors and their descendants (and identify the
strongest predictors) – MH1 and MH2 (see Chapter Ten, Section 10.4.1.1);
2.
that the proposed modes of trauma transmission/family interaction variables are
correlated with the severity or level of psychological impact variables and influential
© Janine Lurie-Beck 2007
212
psychological process variables among descendants of survivors (and identify the strongest
predictors) – MH4 to MH8 (see Chapter Ten, Section 10.4.1.2);
3. that ancestor scores on psychological impact variables are correlated with descendant
scores on psychological impact variables;
4. that the relationship between ancestor and descendant scores on psychological impact
variables is mediated by the proposed modes of trauma transmission/family interaction
variables; and
5. that there is a dissipation/lessening of the effects of Holocaust trauma as evidenced by
improvement in scores on psychological impact variables with each generational removal
from the Holocaust – DH11 (see Chapter Ten, Section 10.4.2).
There are insufficient family dyads/pairs within the sample derived for this study to
enable correlation analyses between ancestor and descendant scores on psychological
impact variables (point three) and thus it is also not possible to test for the mediation of this
relationship by the proposed modes of trauma transmission/family interaction variables
(point four). Specifically, only six survivor-child of survivor pairs/dyads and nine child of
survivor-grandchild of survivor pairs/dyads remain, after screening for missing data and
potential confounding demographic variables (such as the number of ancestors directly or
indirectly affected by the Holocaust) is conducted. However, it is possible to test points
one, two and five and the results of these tests are reported in this chapter.
12.1. – The Role of Influential Psychological Processes in Predicting Severity of
Psychological Impacts
In this section, the ability of the influential psychological processes of coping strategies and
world assumptions to predict survivor and descendant scores on the psychological impact
variables outlined in the model is examined.
12.1.1. – Influence of Coping Strategies
Correlations between the composite scores for maladaptive and adaptive coping strategies
and the psychological impact variables of depression, anxiety, post-traumatic vulnerability,
post-traumatic stress disorder (PTSD) symptoms, positive and negative romantic
attachment dimensions and post-traumatic growth (PTG) are reported in this section. These
correlations serve to address the hypothesis that maladaptive coping strategies will be
positively related to negative psychological symptoms and negatively related to positive
psychological dimensions while adaptive coping strategies will be negatively related to
© Janine Lurie-Beck 2007
213
negative psychological symptoms and positively related to positive psychological
dimensions (MH1).
Table 12.1 presents the correlations between maladaptive and adaptive coping
scores and the psychological impact variables for each of the three generations.
Maladaptive coping is statistically significantly correlated with more variables than
adaptive coping, suggesting it is perhaps more influential in determining the severity of
symptoms experienced than adaptive coping. The four statistically significant correlations
with adaptive coping are all in the anticipated direction (with a negative relationship with
negative variables and a positive relationship with positive variables).
While the
statistically significant correlations with maladaptive coping outnumber the statistically
significant correlations with adaptive coping it is interesting to note that all of the
statistically significant maladaptive coping correlations are with negative impact variables
(with correlations with positive variables not reaching significance). Correlations between
maladaptive coping and depression, anxiety and posttraumatic vulnerability/paranoia are all
of a similar magnitude across the generations. It is interesting to note that there is a much
stronger correlation between maladaptive coping and PTSD symptoms (as measured by the
IES-R) among Holocaust survivors compared to the other negative impact variables
measured.
Table 12.1.
Correlations between coping strategies and psychological impact variables among Holocaust survivors and their
descendants
DASS Depression
DASS Anxiety
PTV
IES-R Total
AAS Positive
Dimensions
AAS Negative
Dimensions
PTGI Total
COPE
Maladaptive
0.39 #
0.37 #
0.40 #
0.63 **
- 0.19
Survivors
(n = 23)
COPE
Adaptive
0.11
0.15
- 0.24
- 0.03
0.44 *
0.05
- 0.26
- 0.02
0.23
Children of Survivors
(n = 68)
COPE
COPE
Maladaptive
Adaptive
0.30 *
- 0.39 **
0.33 **
- 0.11
0.38 **
0.00
Not applicable
- 0.04
0.42 ***
0.35 **
- 0.07
Not applicable
Grandchildren of Survivors
(n = 28)
COPE
COPE
Maladaptive
Adaptive
0.51 **
- 0.16
0.40 *
- 0.02
0.21
- 0.48 **
Not applicable
0.13
- 0.05
- 0.12
0.06
Not applicable
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory
* p < 0.05, ** p < 0.01, # p < 0.10 (denoted only when n < 30).
12.1.2. – Influence of World Assumptions of Benevolence and Meaningfulness
The relationship between world assumptions and psychological impact variables are
considered in this section. Again, correlation analyses are used. The hypothesis, in relation
to world assumptions, is that the strength of belief that the world is benevolent and
© Janine Lurie-Beck 2007
214
meaningful will be negatively related to negative psychological symptoms and positively
related to positive psychological dimensions (MH2).
Correlations for each generation are presented in Table 12.2.
All statistically
significant correlations are in directions consistent with the hypothesis with stronger beliefs
in world benevolence and meaningfulness having a positive influence on scores on
psychological impact variables, with the one exception of the positive correlation between
meaningfulness and anxiety for Holocaust survivors. Overall, it seems that the belief in
world benevolence is more influential than the belief in world meaningfulness. While there
are three statistically significant correlations between impact variables and meaningfulness
for the Holocaust survivor sample, the two statistically significant correlations with
benevolence are of a larger magnitude.
Table 12.2.
Correlations between the assumptions of world benevolence and world meaningfulness and psychological impact
variables among Holocaust survivors and their descendants
DASS Depression
DASS Anxiety
PTV
IES-R Total
AAS Positive Dimensions
AAS Negative Dimensions
PTGI Total
Survivors
(n = 23)
WAS – Benevolence WAS – Meaningfulness
- 0.07
0.30
0.14
0.50 *
- 0.61 **
- 0.36
- 0.03
- 0.06
0.67 **
0.51 *
- 0.02
0.14
0.22
0.17
Children of Survivors
(n = 68)
WAS – Ben. WAS – Mean.
- 0.31 *
- 0.22
0.07
0.10
- 0.44 ***
0.08
Not applicable
0.19
0.05
- 0.25 *
- 0.03
Not applicable
Grandchildren of Survivors
(n = 28)
WAS – Ben. WAS – Mean.
- 0.50 **
- 0.27
- 0.56 **
- 0.30
- 0.56 **
- 0.41 *
Not applicable
0.37 #
0.24
- 0.12
0.06
Not applicable
Notes. WAS = World Assumptions Scale, DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability
Scale, IES-R = Impact of Events Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth
Inventory
* p < 0.05, ** p < 0.01, *** p < 0.001, # p < 0.10 (denoted only when n < 30).
12.1.3. – Summary of the Role of Influential Psychological Processes
A summary of the statistically significant relationships between the influential
psychological processes of coping and world assumptions and the psychological impact
variables is presented in Table 12.3. Looking at these relationships graphically it is clear
that the use of maladaptive coping strategies has a key role in determining the severity of
symptoms experienced across the three generations, and that maladaptive coping strategies
are much more influential or predictive than the use of adaptive coping strategies. The
assumption of world benevolence (that the world is a kind and caring place) is clearly also
more determinant than the assumption of world meaningfulness (that the world is a
predictable and understandable place).
The hypotheses about the roles played by world assumptions and coping strategies
are supported with the further clarification of the type of world assumption (benevolence)
© Janine Lurie-Beck 2007
215
and the type of coping strategies (maladaptive) that are the most strongly predictive. These
results support the presence of world assumptions and coping strategies in the model of the
differential impact of Holocaust trauma as influential psychological processes in predicting
the severity of psychological impact among survivors and descendants.
Table 12.3.
Relationships between influential psychological processes and psychological impact variables
WAS –
WAS –
COPE
COPE Adaptive
Benevolence
Meaningfulness
Maladaptive
Survivors
DASS Depression
Positive
relationship
DASS Anxiety
Positive
Positive
relationship
relationship
PTV
Negative
Negative
Positive
relationship
relationship
relationship
IES-R Total
Positive
relationship
AAS Negative
Dimensions
AAS Positive
Positive
Positive
Positive
Dimensions
relationship
relationship
relationship
PTGI Total
Children of
DASS Depression
Negative
Positive
Survivors
relationship
relationship
DASS Anxiety
Positive
relationship
PTV
Negative
Positive
relationship
relationship
AAS Negative
Negative
Positive
Dimensions
relationship
relationship
AAS Positive
Positive
Dimensions
relationship
Grandchildren DASS Depression
Negative
Positive
of Survivors
relationship
relationship
DASS Anxiety
Negative
Positive
relationship
relationship
PTV
Negative
Negative
Negative
relationship
relationship
relationship
AAS Negative
Dimensions
AAS Positive
Positive
Dimensions
relationship
Notes. WAS = World Assumptions Scale, DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability
Scale, IES-R = Impact of Events Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth
Inventory.
Blank cells indicate the absence of a statistically significant relationship.
12.2. – The Relationship between Posttraumatic Growth and Psychological Impact
Variables
A test of the hypothesis that posttraumatic growth will be positively associated or
correlated with negative symptom levels (MH3) is presented in this section. Table 12.4
© Janine Lurie-Beck 2007
216
reports the correlations between the psychological impact variables and the total score and
subscale scores of the Posttraumatic Growth Inventory. There are very few correlations
that reach even the p < 0.10 level. The correlations with the IES-R which measures PTSD
symptoms and the anxiety subscale of the DASS are consistent with previous research that
found a positive relationship between symptoms and growth.
However, the negative
correlation with the depression subscale of the DASS and the positive correlation with
positive attachment dimensions, as measured by the AAS, suggest the opposite – that
growth and negative symptoms are polar opposites. Also, interestingly, there is no notable
correlation in either direction between growth and vulnerability. Looking at the particular
subscales that are correlated the most with depression and positive attachment dimensions,
the pattern of results obtained is not that surprising. The negative correlation with a
recognition of personal strength and depression scores is certainly predictable. Likewise,
the finding that an increased ability to relate to others would be positively associated with
attachment dimensions, such as comfort with closeness and dependence on others is also
intuitively understandable.
Table 12.4.
Relationships between posttraumatic growth and psychological impact variables among survivors (n = 23)
PTGI Total Score
DASS Depression
DASS Anxiety
PTV
IES-R Total
AAS Positive
Dimensions
AAS Negative
Dimensions
Relating to others
New Possibilities
PTGI Subscales
Personal Strength
Spiritual Change
- 0.30
0.11
- 0.06
0.29
0.35
- 0.33
0.01
- 0.19
0.31
0.49 *
- 0.07
0.26
- 0.10
0.04
0.35
- 0.38 #
- 0.03
0.02
0.19
0.16
0.13
0.37 #
0.15
0.55 **
0.15
Appreciation of
life
- 0.36
- 0.01
0.20
0.37 #
0.09
- 0.13
- 0.20
- 0.13
- 0.05
- 0.19
- 0.11
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory
* p < 0.05, ** p < 0.01, # p < 0.10 (denoted only when n < 30).
12.3. –The Role of the Proposed Modes of Trauma Transmission/Family Interaction
Variables
In this section, the degree to which the family interaction variables (which are the proposed
modes of trauma transmission) are related to/can predict scores on psychological impact
variables and influential psychological processes is considered.
12.3.1. – Influence of Parent-child Attachment Dimensions
It is hypothesised that negative parent-child attachment dimensions, such as the degree of
coldness and ambivalence, will be positively associated with negative psychological
symptoms and negatively associated with positive psychological symptoms, while positive
© Janine Lurie-Beck 2007
217
parent-child attachment dimensions, such as perceived parental warmth, will be negatively
associated with negative psychological symptoms and positively associated with positive
psychological dimensions (MH4).
Both children of survivor/s and grandchildren of
survivor/s perceptions of their parents on parent-child attachment dimensions of warmth,
coldness and ambivalence and how they relate to their scores on psychological impact
variables and influential psychological variables are considered in this section.
Correlations between children of survivor/s perceptions of their survivor parents’
warmth, coldness and ambivalence and children of survivor/s scores on psychological
impact and influential psychological process variables, as displayed in Table 12.5, are all
largely in the expected direction (positive dimensions correlating negatively with negative
variables and positively with positive variables and the opposite for negative dimensions).
Because each parent was considered separately it is possible to compare the strength of
relationships with parent-child attachment between mother and father ratings. To eliminate
confounds with survivor status of parents, these correlations were conducted within the
children of survivor sample subset for whom both parents are survivors. Overall, there are
stronger and more statistically significant relationships with maternal than paternal
dimensions, suggesting Holocaust survivor parent gender (as does parent gender in the
normal population) plays a moderating role.
Table 12.5.
Correlations between children of survivor’s scores on psychological impact and influential psychological process variables
and their perceptions of their survivor parents (among children with two survivor parents only [n = 51])
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
PCS – Warmth
Survivor
Survivor
Mother
Father
PCS – Coldness
Survivor
Survivor
Mother
Father
PCS – Ambivalence
Survivor
Survivor
Mother
Father
- 0.31 *
- 0.37 **
- 0.21
0.52 ***
- 0.33 *
- 0.28 *
- 0.48 ***
- 0.17
0.40 **
- 0.29 *
0.29 *
0.48 ***
0.22
- 0.32 *
0.48 ***
0.32 *
0.48 ***
0.05
-0.18
0.25
0.35 *
0.30 *
0.06
- 0.37 **
0.24
0.10
0.29 *
0.10
- 0.36 *
0.25
- 0.10
0.30 *
0.33 *
0.12
- 0.17
0.17
0.21
0.15
0.12
- 0.15
- 0.32 *
- 0.27
0.17
- 0.02
- 0.15
- 0.10
0.18
- 0.26
- 0.21
- 0.05
0.30 *
- 0.12
- 0.13
- 0.11
Notes. PCS = Parental Care-giving Style Questionnaire, DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic
Vulnerability Scale, AAS = Adult Attachment Scale.
* p < 0.05, ** p < 0.01, *** p < 0.001.
There are far fewer correlations that reach significance for the grandchildren of
survivor/s sample than there are for the children of survivor/s sample, when considering
relationships between perceptions of parent-child attachment and psychological impact
variables (see Table 12.6). However, part of the reason for this is the much smaller sample
© Janine Lurie-Beck 2007
218
size involved (n = 10) meaning much larger correlation co-efficients are required to reach
even the more liberal probability level of 0.10 or less. The correlations that are statistically
significant are for the most part in the area of adult attachment dimensions. Again, it is the
maternal parent-child attachment dimensions that are more strongly related than paternal
dimensions however the size of the correlations are quite large for both parents. All
correlations are in the direction consistent with MH4.
Table 12.6.
Correlations between grandchildren of survivors’ scores on psychological impact and influential psychological process
variables and their perceptions of their child of survivor parents (among those with two child of survivor parents only [n =
10])
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
PCS – Warmth
Child of
Child of
Survivor
Survivor
Mother
Father
PCS – Coldness
Child of
Child of
Survivor
Survivor
Mother
Father
PCS – Ambivalence
Child of
Child of
Survivor
Survivor
Mother
Father
- 0.30
- 0.11
0.15
0.88 **
- 0.92 ***
- 0.05
- 0.15
0.10
0.60 #
- 0.35
0.39
0.04
- 0.09
- 0.89 ***
0.98 ***
- 0.01
- 0.07
- 0.31
-0.74 *
0.62 *
0.47
0.47
0.17
- 0.78 **
0.73 *
0.16
0.37
0.02
- 0.70 *
0.51
- 0.30
- 0.42
0.04
- 0.30
- 0.05
- 0.17
- 0.19
- 0.31
- 0.08
0.57 #
- 0.18
0.08
- 0.28
0.26
0.08
0.45
0.24
0.15
- 0.02
0.15
0.01
0.10
0.28
0.43
Notes. PCS = Parental Care-giving Style Questionnaire, DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic
Vulnerability Scale, AAS = Adult Attachment Scale.
* p < 0.05, ** p < 0.01, *** p < 0.001, # p < 0.10 (denoted only when n < 30).
12.3.2. – Influence of Family Cohesion
It is hypothesised that a curvilinear/U-shaped relationship will exist between negative
psychological symptoms and family cohesion (with very low and very high cohesion
associated with higher symptom levels than mid-range scores) and an that an inverted Ushaped relationship will exist between positive psychological dimensions and family
cohesion (MH5). In this section, the relationships between family cohesion in survivor and
child of survivor families and children’s scores on psychological impact variables and
influential psychological process variables are assessed via correlation analyses to test this
hypothesis. Table 12.7 presents the results of these correlation analyses. As can been seen,
there are numerous statistically significant linear correlations between children of
survivor/s perceptions of cohesion in their family of origin and their scores on both
psychological impact and influential psychological process variables. While two of these
statistically significant correlations are quadratic/curvilinear in nature as predicted by MH5,
it is notable that the remaining four statistically significant correlations are linear and not
curvilinear in nature.
© Janine Lurie-Beck 2007
The four linear correlations agree with the two curvilinear
219
correlations that low cohesion has a negative influence, but disagree that high cohesion also
has a negative influence. It is interesting, however, that it is the two psychopathological
measures of anxiety and depression that appear to be negatively influenced by either very
high or very low cohesion.
Within the grandchildren of survivor/s sample, only two correlations reached
significance with a linear relationship between positive attachment dimensions and
perceptions of family cohesion and a quadratic/curvilinear relationship between the
negative attachment dimension of attachment anxiety and cohesion.
Table 12.7.
Correlations between children and grandchildren of survivor/s scores on impact and influential process variables and their
perceptions of their family of origin cohesion
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
Children of Survivors (n = 66)
Linear
Quadratic
Grandchildren of Survivors (n = 28)
Linear
Quadratic
- 0.40 **
- 0.61 ***
- 0.16
0.50 ***
- 0.35 **
0.47 *
0.68 ***
0.22
0.50
0.35
0.03
0.01
- 0.07
0.49 **
- 0.26
0.23
0.20
0.24
0.49
0.55 **
- 0.10
- 0.41 **
0.28 *
0.15
0.10
0.42
0.28
0.26
0.01
0.03
0.04
0.05
0.02
0.16
0.07
0.12
Notes. FES = Family Environment Scale, DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability
Scale, AAS = Adult Attachment Scale
* p < 0.05, ** p < 0.01, *** p < 0.001.
Figures 12.2 and 12.3 present graphical representations of the three statistically
significant curvilinear relationships with perceptions of family cohesion reported in Table
12.7. It is particularly noteworthy that all of the children of survivor/s who score above
normal on anxiety (as marked on the graph by the horizontal line) perceive their family
cohesion level to be at one of the extremes (either enmeshed or disengaged). The pattern of
extreme cohesion scores accompanying above normal depression scores is not as clearly
curvilinear as is the case for anxiety. It is noteworthy that the very high depression scores
are mostly associated with very low cohesion (or disengagement). Among grandchildren of
survivor/s, the influence of perceptions of family cohesion on attachment anxiety are also
U-shaped and again it is notable that the scores that fall above the normative mean score for
this scale correspond to extreme cohesion scores representing enmeshment or
disengagement.
© Janine Lurie-Beck 2007
220
50.00
DASS Depression Scale Score
DASS Anxiety Scale Score
20.00
15.00
10.00
5.00
0.00
40.00
30.00
20.00
10.00
0.00
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
FES Cohesion Score
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
FES Cohesion Scale Score
Figure 12.2. Scatterplots of children of survivor anxiety and depression with child of survivor perceptions of family
cohesion
Note. The horizontal lines represent the upper most limits of scores within the normal range on the DASS subscales as
provided by the scale authors (S. H. Lovibond & P. F. Lovibond, 1995).
AAS Attachment Anxiety Scale Score
30.00
25.00
20.00
15.00
10.00
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
FES Cohesion Scale Score
Figure 12.3. Scatterplot of grandchildren of survivor/s Negative Attachment Dimension/Attachment Anxiety with
grandchildren of survivor/s Perceptions of Family Cohesion
Note. The horizontal line represents the normative score for the attachment anxiety subscale of the AAS as reported by
Collins and Read (1990).
12.3.3. – Influence of Parental Encouragement of Independence
It is hypothesised that the degree to which parents are encouraging of their children’s
attempts to establish independence, when children are attempting to establish and assert
their autonomy in their late teen and early adult years (MH6), will be negatively associated
with negative psychological symptoms and positively associated with positive
psychological dimensions. This section reports correlation analyses between perceptions of
parental encouragement of independence on children of survivors’ and grandchildren of
survivors’ scores and their scores on psychological impact and influential psychological
process variables.
Perceptions of maternal and paternal encouragement of independence
© Janine Lurie-Beck 2007
221
were measured separately. In order to examine the relative importance of maternal versus
paternal influence without clouding the issue with survivor status of the parent, these
analyses were conducted with children whose parents are both survivors.
The correlations for children of survivor/s and grandchildren of survivor/s with their
perceptions of parental facilitation of autonomy/encouragement of independence are
presented in Table 12.8. As with parent-child attachment dimensions, all correlations with
parental facilitation of independence are in the anticipated direction with negative
symptoms decreasing, and scores on positive dimensions of attachment increasing with
increasing encouragement of independence on the part of parents. This is certainly true
when maternal encouragement is considered; however none of the correlations with
paternal encouragement reached the necessary probability level. Additionally, none of the
influential psychological process variables correlate statistically significantly with either
maternal or paternal encouragement/facilitation of independence.
Table 12.8.
Correlations between child and grandchild of survivor scores on impact and influential process variables and their
perceptions of their survivor and child of survivor parent’s facilitation of independence/fostering of autonomy
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative
Dimensions
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
Children of Survivor perceptions of Survivor Parents
(among children with two survivor parents
[n = 50])
PAQ – Fostering of Autonomy
Mother
Father
Grandchildren of Survivor Perceptions of Child of Survivor
Parents (among those with two child of survivor parents only
[n = 10])
PAQ – Fostering of Autonomy
Mother
Father
- 0.24
- 0.33 *
- 0.13
0.47 **
- 0.25
- 0.07
- 0.23
- 0.12
0.18
- 0.16
- 0.34
- 0.15
0.16
0.79 **
- 0.85 **
- 0.15
- 0.44
- 0.24
0.28
- 0.12
- 0.14
0.23
0.26
0.04
- 0.19
- 0.01
0.13
0.20
0.02
- 0.48
- 0.10
- 0.30
0.15
- 0.03
- 0.40
- 0.18
Notes. PAQ – Parental Attachment Questionnaire, DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic
Vulnerability Scale, AAS = Adult Attachment Scale, WAS = World Assumptions Scale
* p < 0.05, ** p < 0.01.
12.3.4. – Influence of Level of Family Communication
It is hypothesised that general communicativeness within the family unit (MH7) will be
negatively associated with negative psychological symptoms and positively associated with
positive psychological dimensions.
The general level of communicativeness or
expressiveness in children of survivor/s and grandchildren of survivor/s nuclear families is
© Janine Lurie-Beck 2007
222
considered, in this section, in terms of how it relates to their scores on psychological impact
variables.
Table 12.9 reports the results of correlation analyses between the expressiveness
subscale of the Family Environment Scale (FES) and the psychological impact variable
scores of children of survivor/s and grandchildren of survivor/s. There are only three
statistically significant correlations among either children of survivor/s or grandchildren of
survivor/s between family expressiveness and psychological impact variables; however all
three are in the direction hypothesised.
It is interesting that the positive romantic
attachment dimensions, of comfort with closeness and dependence on others, are the
strongest correlations for both generations.
Table 12.9.
Correlations between children and grandchildren of survivors’ scores on impact and influential psychological process
variables and their perceptions of their family of origin expressiveness
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
Children of survivors
(n = 67)
FES – Expressiveness
Grandchildren of survivors
(n = 28)
- 0.18
- 0.27 *
- 0.13
0.39 **
- 0.22
- 0.01
0.03
- 0.08
0.44 *
- 0.27
0.04
0.24 *
0.23
0.15
- 0.11
0.23
0.02
- 0.03
Notes. FES = Family Environment Scale, DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability
Scale, AAS = Adult Attachment Scale, WAS = World Assumptions Scale
* p < 0.05, ** p < 0.01.
12.3.5. – Influence of Communication about the Holocaust
It is hypothesised that negative modes of communicating about the Holocaust, such as
guilt-inducing, indirect and non-verbal communication, will be positively associated with
negative psychological symptoms and negatively associated with positive psychological
symptoms, while positive modes of communicating about the Holocaust, such as frequent,
willing and open discussion will be negatively associated with negative psychological
symptoms and positively associated with positive psychological dimensions (MH8). As
was the case with parent-child attachment dimensions and parental facilitation of
independence, only data for children with both survivor parents are presented here (see
Table 12.10), so that the issue of number of survivor parents does not confound the
analysis, as parental gender is also considered.
© Janine Lurie-Beck 2007
223
Table 12.10.
Correlations between modes of communication about Holocaust experiences and children with two survivor parents’ (n =
51) scores on psychological impact variables.
Guilt Inducing
communication about the
Holocaust
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive
Dimensions
AAS Negative
Dimensions
Influential
Psychological
Processes
WAS –
Benevolence
WAS –
Meaningfulness
COPE Maladaptive
COPE Adaptive
Indirect communication
about the Holocaust
Affective communication
about the Holocaust
Survivor
Mother
Survivor
Father
Survivor
Mother
Survivor
Father
Survivor
Mother
Survivor
Father
0.19
0.04
0.01
- 0.15
0.16
0.28
0.15
- 0.21
0.03
- 0.03
- 0.18
0.03
- 0.00
- 0.15
- 0.00
0.24
0.10
0.13
0.17
- 0.28
- 0.05
0.23
0.19
- 0.35 *
0.16
0.14
- 0.01
- 0.13
0.29 *
- 0.16
0.05
0.01
0.08
- 0.12
0.00
- 0.02
0.14
- 0.17
0.11
- 0.22
0.02
- 0.03
Non Verbal
Presence of
the
Holocaust
Frequent and Willing
communication about
the Holocaust
Survivor
Mother
Survivor
Father
0.38 **
0.40 **
0.39 **
- 0.19
0.08
- 0.10
- 0.19
0.15
- 0.01
- 0.24
0.21
0.23
0.28
0.32 *
0.02
- 0.04
- 0.39 **
- 0.07
- 0.17
0.04
0.04
0.16
- 0.36 *
- 0.12
- 0.05
- 0.00
0.02
- 0.02
0.19
0.31 *
- 0.16
- 0.04
0.03
0.12
- 0.10
- 0.20
- 0.08
0.12
0.10
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale
* p < 0.05, ** p < 0.01.
Given the possible number of results, there are relatively few statistically significant
correlations between modes of communication about the Holocaust and children of
survivor/s scores on psychological impact variables. However, the correlations that do
reach significance are all in the direction predicted. The two communication methods that
statistically significantly relate to psychological impact variables are two negative modes;
namely affect-laden communication and non-verbal presence and they do so in the
anticipated direction (that is they are associated with higher scores on negative variables).
Again, it is the maternal line that is more predictive than the paternal line. The positive
mode of communication (frequent and willing communication) is not statistically
significantly correlated with any of the psychological impact variables which adds further
support to the notion that it is not so much the content of the communication that is
potentially damaging but the way in which it is conveyed.
12.3.6. – Summary of the Influence of Family Interaction Variables
Tables 12.11 and 12.12 provide a visual summary of the relationships between the family
interaction variables/proposed modes of trauma transmission and children of survivor/s and
grandchildren of survivor/s scores on psychological impact and influential psychological
process variables. All of the relationships between family interaction/intergenerational
transmission variables and children of survivor/s and grandchildren of survivor/s scores on
© Janine Lurie-Beck 2007
224
psychological impact variables are in the hypothesised directions. Generally there are more
statistically significant correlations for maternal interaction variables when parental gender
is considered separately.
Among the children of survivor/s it is the parent-child attachment, family cohesion
and communication about the Holocaust variables that are related to the largest number of
psychological impact variables from the model.
Parental warmth is associated with
positive outcomes, while parental coldness and ambivalence are associated with negative
outcomes. The more cohesive a family is perceived to be, the more positive the effects are
for the children; however, for both depression and anxiety the u-shaped relationships
suggest that too much cohesion (enmeshment) is as detrimental as not enough
(disengagement). The styles of communication most predictive are the negative modes of
affective communication and the silence of a non-verbal unspoken presence of the
Holocaust which both have negative influences on psychological functioning.
Among the grandchildren of survivor/s, parent-child attachment is again the factor
that is linked to the most number of variables; however family cohesion and parental
fostering of autonomy are also implicated more than once. The influence of maternal
attachment behaviour is again much stronger than paternal attachment behaviour. Notably
all the statistically significant correlations for this generation are with the romantic
attachment dimensions and not with any of the psychopathology measures.
© Janine Lurie-Beck 2007
225
Table 12.11.
Statistically significant relationships between proposed modes of trauma transmission/family interaction variables and psychological impact variables and influential psychological
processes among children of survivors
Parental Care-giving Style
Parental Attachment
Family
Family
Holocaust Communication Questionnaire
Questionnaire
Questionnaire –
Environment
Environment
Fostering of Autonomy
Scale –
Scale –
Cohesion
Expressiveness
Maternal
Paternal
Maternal
Paternal
Maternal
Paternal
Non-Verbal
Presence
Psychological
Impacts
DASS
Negative
Negative
Negative
U-shaped
Negative
Positive
Depression
relationship with relationship with
relationship
relationship
relationship
relationship
warmth
warmth
Positive
Positive
relationship with relationship with
coldness and
coldness and
ambivalence
ambivalence
DASS Anxiety
Negative
Negative
U-shaped
Positive
relationship with relationship with
relationship
relationship
warmth
warmth
Positive
Positive
relationship with relationship with
coldness and
coldness
ambivalence
PTV
Positive
relationship
AAS Negative
Negative
Negative
Negative
Positive relationship
Positive
Dimensions
relationship with relationship with
relationship
with affective
relationship
warmth
warmth
communication about
Positive
Positive
the Holocaust
relationship with relationship with
coldness and
coldness and
ambivalence
ambivalence
© Janine Lurie-Beck 2007
226
Parental Care-giving Style
Questionnaire
Maternal
AAS Positive
Dimensions
Influential
Psychological
Processes
COPE
Maladaptive
COPE Adaptive
WAS –
Benevolence
WAS –
Meaningfulness
Positive
relationship with
warmth
Negative
relationship with
coldness and
ambivalence
Paternal
Positive
relationship with
warmth
Negative
relationship with
ambivalence
Parental Attachment
Questionnaire –
Fostering of Autonomy
Maternal
Positive
relationship
Paternal
Family
Environment
Scale –
Cohesion
Positive
relationship
Family
Environment
Scale –
Expressiveness
Paternal
Negative
relationship with
affective
communication
about the
Holocaust
Non-Verbal
Presence
Positive relationship
with affective
communication about
the Holocaust
Negative
relationship
Positive
relationship
Maternal
Positive
relationship
Positive
relationship with
ambivalence
Positive
relationship with
warmth
Positive
relationship with
warmth Negative
relationship with
coldness
Holocaust Communication Questionnaire
Positive
relationship
Negative relationship
with affective
communication about
the Holocaust
Negative relationship
with affective
communication about
the Holocaust
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment Scale, WAS = World Assumptions Scale
Blank cells denote an absence of statistically significant relationships.
© Janine Lurie-Beck 2007
227
Table 12.12.
Statistically significant relationships between proposed modes of trauma transmission/family interaction variables and
psychological impact variables and influential psychological processes among grandchildren of survivors
Parental Care-giving Style
Parental Attachment
Family
Family
Questionnaire
Questionnaire –
Environment
Environment
Fostering of Autonomy
Scale –
Scale –
Cohesion
Expressiveness
Maternal
Paternal
Maternal
Paternal
Psychological
Impacts
DASS
Depression
DASS Anxiety
PTV
AAS Negative
Positive
Negative
U-shaped
Dimensions
relationship with
relationship
relationship
warmth
Negative
relationship with
coldness and
ambivalence
AAS Positive
Positive
Positive
Positive
Positive
Positive
Dimensions
relationship with
relationship relationship
relationship
relationship
warmth
with warmth
Negative
relationship with
coldness and
ambivalence
Influential
Psychological
Processes
COPE
Maladaptive
COPE Adaptive Negative
relationship with
warmth
Positive
relationship with
coldness
WAS –
Positive
Benevolence
relationship
WAS –
Meaningfulness
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
Blank cells indicate the absence of a statistically significant relationship.
12.4. – Intergenerational Differences
In this section, the hypothesis that there will be a dissipation of the impact of the Holocaust
with each generational separation from the Holocaust (DH11) will be examined.
12.4.1. – Intergenerational Differences on Psychological Impact Variables
Tables 12.13 and 12.14 present comparisons between the three generations of survivors and
normative data on the psychological impact variables from the model being tested in the
© Janine Lurie-Beck 2007
228
current thesis. Table 12.13 reports mean scores on these scales while Table 12.14 reports
the percentages of each generation that fall within the normal range as defined by measure
authors.
Table 12.13.
Intergenerational differences in scores on psychological impact variables (including statistically significant differences)
Impact Variables
DASS Depression
DASS Anxiety
PTV Scale
AAS Positive
Dimensions
AAS Negative
Dimensions *
Survivors
(n = 23)
Children of survivors
(n = 69)
Grandchildren of
survivors
(n = 27)
Top Score in
Normal Range
Normative
Data
6.13 (6.32)
4.87 (6.04)
10.83 (4.28)
38.29 (8.20)
7.50 (9.10)
3.43 (3.96)
10.66 (4.73)
40.32 (9.66)
6.22 (7.84)
5.19 (5.99)
10.48 (4.37)
40.11 (10.36)
9.00
7.00
-
5.55
3.56
8.69
39.5
12.00 (3.79)
14.71 (6.26)
16.52 (5.44)
-
16.2
Notes. AAS Negative Dimensions = Kruskal-Wallis χ (2) = 7.38, p < 0.05
Normative data for DASS based on a sample of 1,771 members of the general adult population (Crawford & Henry,
2003). Normative data for the PTV scale based on a convenience sample of 686 undergraduate students and adults from
the general population (Shillace, 1994). Normative data (provided to only one decimal point) for the AAS based on a
sample of 406 undergraduate students (Collins & Read, 1990).
* p < 0.05
2
Firstly, in relation to the psychopathology measures of depression, anxiety, and
posttraumatic vulnerability/paranoia there are some counter-intuitive findings, but some
that, despite this, are consistent with meta-analytic findings.
Scores on depression and
anxiety certainly do not follow the predicted decrease with each generational separation
from the Holocaust. In fact, it is the grandchildren that score the highest on anxiety. While
none of these differences are significant, it is interesting to note that the result of
grandchildren scoring higher than children of survivors is consistent with the seemingly
incongruous findings of the meta-analysis conducted for the current thesis which are of a
similar pattern.
The only variable to follow the predicted pattern is posttraumatic
vulnerability/paranoia which clearly decreases through the generations. However, all three
generations almost uniformly score above the normative mean scores for depression,
anxiety and posttraumatic vulnerability shown in Table 12.13. Attempts were made to find
more recent normative data for the PTV scale (or at least data collected after the September
11 terrorist attacks in 2001) but were unsuccessful. However, the normative data obtained
for the DASS is relatively recent and it is interesting to note that the survivor and
descendant groups score higher than the norms even within the current world political
climate. Turning to the romantic attachment dimensions tabulated in Table 12.13, children
of survivor/s and grandchildren of survivor/s score higher on positive dimensions than the
survivors but not statistically significantly so.
© Janine Lurie-Beck 2007
The statistically significant difference
229
between the generations for the negative attachment dimension of attachment anxiety is
curiously in the exact opposite direction to what is predicted in the hypothesis, with
attachment anxiety increasing rather than decreasing through the generations. However,
levels seen for the grandchildren of survivor/s, which is the highest of the three generations,
is at a similar level to the normative score (which is based on a similarly aged group).
Another way of considering the psychological impact variables across the
generations is to look at the percentage of each generation group that score within the
normal range (or are classified with a secure attachment for the attachment variable). To be
consistent with the hypothesis, it would be expected that with each generational removal
from the Holocaust, the percentage classified within normal range or securely attached will
increase. An examination of Table 12.14 demonstrates that this is the case for depression
but not for anxiety (which is consistent with the mean score results presented in Table
12.13). The percentages of each generation that fall within the normal range for the DASS
are still lower than those found for the general population. Also there is a decrease in the
proportion of each generation that can be classified as secure.
Table 12.14.
Intergenerational differences in the percentage of samples scoring within normal range of tests
DASS Depression
DASS Anxiety
Attachment
Survivors
(n = 23)
70%
78%
61%
Children of survivors
(n = 69)
75%
86%
50%
Grandchildren of survivors
(n = 27)
78%
78%
48%
General
Population
82%
94%
56%
Note. DASS = Depression Anxiety Stress Scales
For attachment the percentage represents the percentage classified with a secure attachment style. Percentage of
normal population classified as within normal range for the DASS derived from normative data published in Crawford and
Henry (2003) based on a sample of 1,771 members of the general adult population. Percentage of normal population
generally classified with a secure attachment style based on Hazan and Shaver (1987).
12.4.2. – Intergenerational Differences on Influential Psychological Processes
In this section, intergenerational differences on the two influential psychological process
variables of coping and world assumptions are assessed.
This involves comparisons
between the generations on scores on the COPE and the benevolence and meaningfulness
scales of the WAS (see Table 12.15).
© Janine Lurie-Beck 2007
230
Table 12.15.
Intergenerational differences in scores on influential psychological process variables (including statistically significant
differences)
COPE Scales
• Maladaptive
• Adaptive
WAS Scales
• Assumption of world benevolence
• Assumption of world meaningfulness
Survivors
(n = 23) a
Children of survivors
(n = 69) b
Grandchildren of survivors
(n = 27) c
42.22 (8.65)
101.61 (15.88)
40.52 (7.52)
95.12 (16.90)
40.66 (8.34)
90.96 (17.52)
30.04 (6.92) b
32.08 (9.15)c
34.38 (7.00) a
33.08 (7.41) c
34.55 (7.67)
37.46 (10.77) a b
Note. WAS = World Assumptions Scale. Lettered superscripts indicate which groups differ statistically significantly
according to Tukey post-hoc analyses.
There are no statistically significant differences between the generations in the level
of usage of maladaptive or adaptive coping strategies. However, it is interesting to note
that the use of adaptive coping strategies decreases with each generational removal from
the Holocaust. Usage levels of maladaptive coping strategies are also slightly higher
among survivors, but their usage levels are much more uniform across the three
generations.
In relation to world assumptions, the results are mostly consistent with the
dissipation hypothesis, with improvements being noted with generational separation from
the Holocaust. Certainly, survivor beliefs in world benevolence and meaningfulness are
statistically significantly lower than at least one of the descendant groups. It is interesting
to note that a noticeable jump in the belief of world benevolence occurs for the children of
survivors, whereas the increase in the belief in world meaningfulness occurs only at the
grandchildren of survivor stage. Could it be that it takes further separation from the events
of the Holocaust for beliefs that the world is fair and just to return to more normal levels?
Unfortunately, no normative data with a non-traumatised sample could be located for the
World Assumptions Scale. It would be of interest to note how survivor and descendant
scores on this measure compare to a general population sample.
12.3.3. – Intergenerational Differences on Perceptions of Family Interaction
In this section, differences between how children of survivors view their parents and how
grandchildren of survivors view their parents are examined. To be consistent with the
hypothesis that there will be a gradual improvement with each generational separation from
the Holocaust (DH11), it is hypothesised that child of survivor parents will be viewed more
positively than survivor parents. The results reported in this section are supportive of this
hypothesis.
© Janine Lurie-Beck 2007
231
Table 12.16 presents perceptions of survivors and children of survivors as parents in
terms of parent-child attachment dimensions of warmth, coldness and ambivalence as well
as the degree to which they encouraged/facilitated their children’s independence. Overall,
children of survivors are perceived more positively by their children than survivors in terms
of these parenting dimensions. Three of these differences reached statistical significance.
Survivor mothers are rated as having been less warm (t (75) = 2.74, p < 0.01) than child of
survivor mothers, while survivor fathers are rated as having been less warm (t (71) = 2.51,
p < 0.05) and less facilitating of their children’s independence (t (28.98) = 3.89, p < 0.01)
than child of survivor fathers.
Table 12.16.
Mean differences in ratings of survivor parents versus child of survivor parents on parent-child attachment dimensions
and parental facilitation of independence
Mother
• PCS – Warm **
• PCS – Cold
• PCS – Ambivalent
• PAQ – Fostering of autonomy
Father
• PCS – Warm *
• PCS – Cold
• PCS – Ambivalent
• PAQ – Fostering of autonomy **
Perception of Survivor Parents
n = 55
8.38 (5.28)
2.95 (4.03)
5.15 (4.89)
43.47 (14.96)
n = 61
8.52 (5.05)
3.70 (4.07)
5.38 (4.38)
45.30 (14.07)
Perception of Child of Survivor Parents
n = 22
11.89 (4.51)
1.41 (3.00)
3.86 (5.15)
48.86 (14.49)
n = 12
12.42 (4.10)
1.67 (3.63)
4.33 (4.72)
54.21 (7.84)
Notes. PCS = Parental Care-giving Style questionnaire, PAQ = Parental Attachment Questionnaire
* p < 0.05, ** p < 0.01.
On the family cohesion and expressiveness scales of the Family Environment Scale (FES),
the families of origin of the children of survivor/s (where the parents are the survivors) are
rated more negatively than the grandchildren of survivor/s families of origin. Specifically,
children of survivor/s rate their families as less cohesive and statistically significantly less
expressive or communicative then grandchildren of survivor/s rate their families (see Table
12.17).
Table 12.17.
Mean differences in ratings of survivor versus child of survivor families on family cohesion and expressiveness
Cohesion
Expressiveness **
Survivor Parent
families (n = 67)
42.04 (22.57)
36.23 (18.15)
Child of Survivor Parent
Families (n = 28)
48.18 (21.82)
50.07 (15.65)
Normative Score for
General Population
49.88
49.70
Normative Score for Distressed
Families
38.24
44.60
Notes. FES = Family Environment Scale. Normative data sourced from Moos (1974). “Distressed families” include
families with issues surrounding psychiatric diagnosis, children in crisis, alcohol abuse, and criminal records.
** p < 0.01.
In relation to normative scores for the FES, cohesion within survivor parent families
is rated notably lower than for the general population, but higher than normative data
© Janine Lurie-Beck 2007
232
collected from distressed families. Cohesion within families with at least one parent who is
a child of survivors is very close to the normative score for the general population. Among
survivor families expressiveness is rated very much lower than both the normative score for
the general population and for distressed families. In stark contrast to this is the level of
expressiveness in the families of the next generation, which is very close but higher than
the normative score for the general population (and significantly higher than that noted in
the survivor parent families).
12.5. – Summary and Conclusions
The model of the differential impact of Holocaust trauma across three generations has been
tested segmentally in this chapter. To test the model fully it would be necessary to:
1. Establish that the influential psychological processes could predict the severity or level
of psychological impact variables among survivors and their descendants.
2. Establish that the proposed modes of trauma transmission/family interaction variables
could predict the severity or level of psychological impact variables and influential
psychological process variables among descendants of survivors.
3.
Establish that ancestor scores on psychological impact variables could predict
descendant scores on psychological impact variables.
4. Establish that the relationship between ancestor and descendant scores on psychological
impact variables is mediated by the proposed modes of trauma transmission/family
interaction variables.
5. Establish the dissipation/lessening of the effects of Holocaust trauma by improvement in
scores on psychological impact variables with each generational removal from the
Holocaust.
In relation to point one, it was established that the influential psychological process
variables in the proposed model, coping strategies and world assumptions, could both be
used to predict survivor and descendant scores on psychological impact variables. It was
further established that maladaptive coping strategies are more strongly related to
psychological variables than adaptive coping strategies and that the belief in world
benevolence is more strongly related to them than the belief in world meaningfulness.
Point two seeks to quantify the strength of the relationship between the family
interaction variables (which are the proposed modes of trauma transmission) and the
psychological impact and influential psychological process variables. To this end, it was
determined that all five of the family interaction variables, namely parent-child attachment,
© Janine Lurie-Beck 2007
233
parental fostering of autonomy, family cohesion, family expressiveness and communication
about the Holocaust, are related to the varying numbers of the psychological impact and
influential psychological process variables. All statistically significant relationships are in
the directions hypothesised.
Based on the current sample, it appears that the parent-child attachment dimensions
of parental warmth, coldness and ambivalence are certainly among the most predictive of
the family interaction variables. Maternal attachment behaviour is much more strongly
related to descendant outcomes than paternal attachment behaviour. Family cohesion was
the next most strongly predictive variable for children of survivor/s and grandchildren of
survivor/s outcomes.
In terms of communication about the Holocaust on the part of survivors parents to
their children, a very strong pattern emerged with the negative communication modes of
affective communication (where Holocaust experiences are communicated in a heavily
negative-affect laden way) and non-verbal communication (which implies a non-verbal
ambiguous presence of the Holocaust in the home) being clearly the most related to
children of survivor/s scores on psychological impact variables. It is interesting to note the
lack of statistically significant relationships with the positive frequent and willing
communication about the Holocaust (given the extensive anecdotal literature attesting to its
positive influence).
In addition, the fact that there are no statistically significant
relationships with guilt-inducing communication about the Holocaust (which is also affectladen) is puzzling given the strong relationships with affective communication. Affective
communication refers to the affective state of the survivor when relaying their Holocaust
experiences to their children. Perhaps when a survivor had strong affective reactions when
talking about their experiences (either anger or noticeable grief and sadness), their children
were deeply affected by the sight of their visibly affected survivor parent. On the other
hand guilt-inducing communication refers to situations where for example a child may have
been disobedient and the parent referred to their Holocaust experiences while disciplining
the child (for example “How can you do this to me after all I went through”). While both
styles of communication are laden with affect, perhaps it is the kind of communication
tapped by the affective communication subscale that reflects parental vulnerability which in
turn could be argued to be the most detrimental to impressionable children.
Communication about the Holocaust is a more powerful predictor than general
© Janine Lurie-Beck 2007
234
communication among family members: this hints at the important role that the knowledge
of parental Holocaust survivorship has for the children of survivor/s generation.
The disparate magnitudes and numbers of correlations with parental fostering of
autonomy and family cohesion are also interesting to note. Given the large amount of
discourse in the literature about the difficulties children of survivor/s face/faced during the
separation/individuation phase, it was expected that parental fostering of autonomy would
be a strong predictor of children of survivor/s scores on psychological impact variables.
Problems with separation/individuation are seen to be symptomatic of problematic family
cohesion and the two are certainly related (both theoretically and statistically in the current
study data set). However, family cohesion is more strongly and more frequently related to
psychological impact scores than parental fostering of autonomy. Partial correlation
analyses with both family cohesion and parental fostering of autonomy, alternately
controlling for the influence of the other, led to the same results: family cohesion remained
statistically significantly related to all of the variables it did before partialling out parental
fostering of autonomy and parental fostering of autonomy was no longer statistically
significantly correlated to anything once family cohesion was partialled out. This suggests
that while separation-individuation problems (as measured in this study by parental
fostering of autonomy) are symptomatic of family cohesion problems, it is the family
cohesion problems overall and not those specifically related to the separation-individuation
phase that are the more determinant of psychological symptom and dimension scores.
Figure 12.4 provides a visual summary of the relative importance of the predictive
variables from the model of the differential impact of Holocaust trauma across three
generations. As can be seen, the ordering of the importance of influential psychological
process variables and intergenerational transmission/family interaction variables is exactly
mirrored across the three generations. The ordering is determined based on the number of
statistically significant correlations as well as the magnitude of correlations reported
throughout this chapter.
Among the influential psychological process variables, it is the use of maladaptive
coping strategies that is the most strongly predictive of scores on psychological impact
variables. The two world assumptions dimensions of world benevolence (that the world is
a kind and caring place) and meaningfulness (that the world is a predictable and fair place)
are more strongly related to psychological impact scores than the use of adaptive coping
strategies.
© Janine Lurie-Beck 2007
235
Among the transmission variables, parent-child attachment clearly is a very
powerful force in determining the psychological health of children, followed by family
cohesion. It is interesting to note that for the children of survivors, parental communication
about the Holocaust is ranked third, above parental encouragement of independence and
more tellingly above general family communication. So within a survivor family, the
extent to which a survivor communicated with their children about their Holocaust
experiences and the way in which they did this is more important than how communicative
they were on any other aspect of life.
It should be noted by the reader that the current author acknowledges that the rank
ordering of the transmission modes/family interaction variables via correlations is not as
ideal as it would be via multiple regressions.
However, the sample sizes involved
precluded the valid conduct of such analyses (as evidenced by uninterpretable and
nonsensical output derived when these analyses were attempted).
3rd Generation
(Grandchildren of
Survivors)
2nd Generation
(Children of Survivors)
1st Generation
(Survivors)
Psychological Impact
Variables
Depression
Anxiety
Paranoia
PTSD Symptoms
Romantic Attachment
Dimensions
Post-traumatic growth
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
Ranking of Influential
Psychological Processes
1.

3.
4.
1.
2.
3.

Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
2.
4.
1.

2.
3.
4.
Maladaptive coping
strategies
Assumption of World
Benevolence
Assumption of World
Meaningfulness
Adaptive coping Strategies
Maladaptive coping
strategies
Assumption of World
Benevolence
Assumption of World
Meaningfulness
Adaptive coping Strategies
Maladaptive coping
strategies
Assumption of World
Benevolence
Assumption of World
Meaningfulness
Adaptive coping strategies
Ranking of Modes of
Intergenerational Transmission of
Trauma
1.
2.
3.
4.
5.
1.
2.
3.
4.
Parent-Child Attachment
(especially maternal)
Family Cohesion
Communication about
Holocaust experiences
(specifically via affective or
non-verbal modes)
Encouragement of
Independence (maternal)
General Family Communication
Parent-Child Attachment
(especially maternal)
Family Cohesion
Encouragement of
Independence (maternal)
General Family Communication
Figure 12.4. Ranking (from most important to least important) of Influential Psychological Processes and Family
Interaction Variables/Proposed Modes of Trauma Transmission in terms of their relative importance in predicting scores
on Psychological Impact Variables
© Janine Lurie-Beck 2007
236
As was mentioned at the outset of this chapter, data restrictions precluded the
assessment of points three and four (the relationship between ancestor and descendant
scores on psychological impact variables and the mediation of this relationship by family
interaction variables). However, there is substantial evidence from the literature with other
populations to support these relationships. As was discussed in Chapter Three, higher rates
of psychopathology among the children of people with psychopathological symptoms or
disorders has been established in the general population. The mediation of parent and child
symptom levels by parent-child attachment has been established in at least two studies (see
Section 3.1.2 of Chapter Three). The potential mediatory role of the remaining family
interaction processes of family cohesion, parental fostering of autonomy, general
communication and also communication about Holocaust experiences (specifically for the
Holocaust population) have yet to be sufficiently tested.
The hypothesised dissipation of the impact of the Holocaust with generational
separation from the Holocaust, as stated in point five, has been tested with the current study
data.
There is support for the dissipation hypothesis in relation to post-traumatic
vulnerability/paranoia and the world assumptions of benevolence and meaningfulness.
However, the findings in relation to depression and anxiety certainly do not fit into the neat
linearly decreasing pattern that was hypothesised. The fact that the upturn in anxiety levels
among grandchildren of survivor/s found with the current study data replicates the pattern
found in the meta-analyses, also reported in the current thesis, “begs the question” of
whether this upturn is a statistical anomaly or reflective of a real pattern. If it is, then the
question of why grandchildren of survivor/s are more anxious than their parents needs to be
asked. It is acknowledged that there are many factors, both personal to the grandchildren of
survivor/s as well as at a community, national and global level that could lead to the
younger generation feeling more anxious. Clearly, the reason for the disparity in anxiety
levels between the generations is an area that requires further investigation. However,
despite the differences between generations on psychological impact variables, survivors
and their descendants still score less favourably than the general population (as represented
by normative data for the measures used in the current thesis).
An improvement in family interaction patterns was clearly noted to correspond with
successive generational removal from direct Holocaust trauma. Perceptions of child of
survivor parents were certainly improved in comparison to perceptions of survivor parents
in relation to attachment patterns and the degree to which they were perceived to encourage
237
© Janine Lurie-Beck 2007
their children’s independence. Marked improvements were also seen in the general family
atmosphere variables of cohesion and expressiveness.
Overall, no contradictory results were obtained in relation to the model of the
differential impact of Holocaust trauma that has been tested in this chapter. The model has
been further refined by prioritising/establishing the relative importance of the proposed
influential psychological processes and proposed modes of trauma transmission. Chapter
Thirteen goes on to test the expanded version of this model which incorporates the
demographic and situational variables which are hypothesised to moderate scores on
psychological impact variables and the flow of the relationships in the model.
© Janine Lurie-Beck 2007
238
Chapter Thirteen – Empirical Assessment of the Moderating Role of Holocaust Survivor
Demographic Variables
Chapter Twelve reported on analyses designed to test the model of the differential impact
of Holocaust trauma across three generations in terms of the three classes of psychological
variables, namely psychological impact, influential psychological process and family
interaction/trauma transmission mode variables. In this chapter, the numerous demographic
variables that have been discussed in the literature (see Chapter Three) and hypothesised to
play a moderating role on model variables, as well as relationships between model variables
(see Chapter Ten), are analysed in terms of their function in the model.
The specific demographic variables to be tested are presented in the representation
of the model in Figure 13.1. They have been bolded. A number of demographic variables
cannot be tested by the current study/sample. These variables are the survivor
demographics of reason for persecution (as no non-Jewish survivors were obtained), as well
as time lapse since the Holocaust (as this is a longitudinal variable which therefore requires
a repeated measures/longitudinal research design).
The descendants of survivor samples (both children and grandchildren) are analysed
both in terms of demographic variables relevant to their generation, as well as demographic
variables related to their ancestors. Specifically, children of survivors are analysed in terms
of both their survivor parent demographic variables as well as the demographics intrinsic to
their own generation, and the grandchildren of survivors are analysed in terms of both their
grandparents and parents demographics as well as those relating to their own generation.
The influence of a number of ancestral demographic variables cannot be tested for all three
generations because of lack of data or prohibitive sample sizes. Such variables are noted at
the relevant juncture in the chapter.
At the end of this chapter, a rank ordering of survivor demographics in terms of the
strength of their relationship to survivor and descendant scores on psychological impact
variables is produced. Thus, as well as assessing each demographic variable in turn, this
chapter provides an idea of the relative importance of each demographic variable in order to
further refine the model of the differential impact of Holocaust trauma.
© Janine Lurie-Beck 2007
239
2nd Generation
(Children of Survivors)
1st Generation
(Survivors)
Psychological Impacts of
the Holocaust
Depression
Anxiety
Paranoia
PTSD symptoms
Romantic Attachment
Dimensions
• Post-traumatic Growth
•
•
•
•
•
•
•
•
•
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
Ranking of Influential Psychological
Processes
Ranking of Modes of Intergenerational
Transmission of Trauma
1.
2.
3.
4.
Maladaptive coping strategies
Assumption of World Benevolence
Assumption of World Meaningfulness
Adaptive coping Strategies
•
•
•
•
•
•
•
•
•
1.
2.
3.
4.
Maladaptive coping strategies
Assumption of World Benevolence
Assumption of World Meaningfulness
Adaptive coping Strategies
•
•
•
•
•
•
•
•
•
1.
2.
3.
3rd Generation
(Grand-children of Survivors)
4.
•
•
•
•
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
Demographic Moderators
Holocaust Survivor Generation
1.
2.
3.
4.
Maladaptive coping strategies
Assumption of World Benevolence
Assumption of World Meaningfulness
Adaptive coping Strategies
5.
1.
2.
3.
4.
Parent-Child Attachment (especially
maternal)
Family Cohesion
Communication about Holocaust
experiences (specifically via affective or
non-verbal modes)
Encouragement of Independence
(maternal)
General Family Communication
Parent-Child Attachment
(especially maternal)
Family Cohesion
Encouragement of Independence
(maternal)
General Family Communication
•
•
•
•
•
•
•
•
•
Age during the Holocaust
Time lapse since the Holocaust – Unable to test
Gender
Type/nature of Holocaust experiences
Reason for persecution – Unable to test
Loss of family
Country of origin
Post-war settlement location
Length of time before resettlement/time spent in
displaced persons camps
Age during the Holocaust
Time lapse since the Holocaust – Unable to test
Gender
Type/nature of Holocaust experiences
Reason for persecution – Unable to test
Loss of family
Country of origin
Post-war settlement location
Length of time before resettlement/time spent in
displaced persons camps
Age during the Holocaust
Time lapse since the Holocaust – Unable to test
Gender
Type/nature of Holocaust experiences
Reason for persecution – Unable to test
Loss of family
Country of origin
Post-war settlement location
Length of time before resettlement/time spent in
displaced persons camps
Figure 13.1. Addition of Holocaust Survivor Descendant Demographic Moderators to Testing Model of the differential impact of Holocaust Trauma across Three Generations
© Janine Lurie-Beck 2007
240
13.1. – Demographic Variable Inter-relationships
Where statistically significant relationships between demographic variables exist, it is
possible that they may confound the analysis of relationships between demographic
variables and variables in the model. Therefore, the inter-relationships among survivor
demographic variables are examined first and the results of this process are reported in this
section. Where statistically significant relationships exist, they serve to inform the use of
controlled statistical analysis, such as ANCOVA or partial correlations, when considering
relationships between demographic variables and model variables.
There are a number of notable relationships between the survivor demographic
variables that may lead to misleading results when looking at their impact univariately. A
survivor’s age during the Holocaust (as operationalised by their age in 1945) differs
statistically significantly on a number of variables. Specifically, these are the nature of
their Holocaust experiences, the region they currently reside in and whether they are a sole
survivor of their family.
Within the nature of Holocaust experience categories, those who escaped prior to
1945 and did not experience the full gamut of possible traumas were aged on average 15.67
years in 1945 (SD = 4.37, n = 6), those who spent time in a camp were aged 23.08 years
(SD = 7.29, n = 12) and those who had other Holocaust experiences such as living in
hiding were aged only 10.78 years (SD = 8.57, n = 9) on average. Specifically, it is the
camp survivors who were statistically significantly older in 1945 than the survivors with
other experiences (F (2, 24) = 7.57, p < 0.01). This result is not surprising given the low
likelihood of survival of children in camps and much higher likelihood of children
surviving in hiding. Survivors who are the sole survivor of their family (M = 23.40, SD =
6.58, n = 5) were much older in 1945 than those who had surviving family members (M =
15.95, SD = 8.91, n = 22; (t (25) = 1.75, p = 0.092).
The final notable difference relating to age is that of the survivor’s current region of
residence. Specifically the three participants who remained in Europe (M = 7.33, SD =
4.04) were much younger than those who immigrated to other regions of the world
(America M = 14.63, SD = 4.50, n = 8; Australia/New Zealand M = 20.33, SD = 9.84, n =
15; Israel 24, n = 1; F (3, 23) = 2.70, p = 0.070). This result is also reflected in the
correlation between age and the number of years spent in Europe after the war (r = - 0.35, p
= 0.083).
© Janine Lurie-Beck 2007
241
The Holocaust experience categories of escapees, camp survivors and noncamp/hiding survivors differ in terms of the proportion who spent at least part of the war
separated from family members and whether they are the sole survivor of their family. This
is not surprising, as separation from family was an intrinsic part of the process of ending up
in a camp, and those in hiding were more likely to have been with other family members.
Specifically, the result for experience type by sole survivor status is χ 2 (2) = 7.67, p <
0.05. All sole survivors in the survivor sample are camp survivors and all survivors in the
non-camp/hiding and escapee groups had at least one surviving family member. The χ 2 for
experience type by whether the survivor ever spent time alone without a family member
also reaches the less than 0.10 threshold ( χ 2 (2) = 4.81, p = 0.090).
The final category of relationships between demographic variables that is worthy of
mention is associations with the time a survivor spent in Europe after the war and before
they were able to immigrate to their chosen country away from Europe. Camp survivors
(M = 6.17, SD = 3.97, n = 12) and non-camp/hiding survivors (M = 8.20, SD = 9.58, n =
5) stayed in Europe for a much larger number of years than those who managed to escape
persecution prior to 1945 (M = 0.67, SD = 1.63, n = 6). This group not only made an early
escape from persecution but a rapid escape from the entire continent. Survivors who
remained in Europe until the end of the war were then faced with huge competition among
millions of refugees for the limited immigration places available (F (2, 20) = 3.24, p =
0.060). Somewhat consistent with the finding that camp survivors were more often alone
than non-camp survivors, the number of years spent in Europe after the war by survivors
who spent some time alone during the war is higher than those who always had a family
member with them (M = 8.00, SD = 7.45, n = 9 versus M = 2.89, SD = 3.92, n = 9 t (16) =
1.82, p = 0.087).
The relationship between survivor parent variables was also assessed to check for
the continuation of possible confounds noted for the survivor sample within the child of
survivor sample.
None of the statistically significant inter-relationships between
demographic variables noted for the survivor sample are repeated for the survivor parents
of the children of survivor sample.
13.2. – Moderating Influence of Holocaust Survivor Demographics
In this section, the possible moderating influence of Holocaust survivor demographic and
situational variables on the variables and relationships in the model of the differential
© Janine Lurie-Beck 2007
242
impact of Holocaust trauma is addressed. Each potential moderator is discussed with
respect to its effect on survivor, child and grandchild scores on psychological impact and
influential psychological process variables as well as on the family interaction/proposed
modes of trauma transmission variables.
Co-efficients of determination (r2) and eta-
squareds ( η 2 ) are cited to provide a quantitative method of determining a rank ordering of
the demographic variables in terms of their influence on variables in the model. Both of
these statistics represent the proportion of variance in the dependent variables (model
variables) accounted for by the independent variables (demographic variables): r2 for
correlation analyses and η 2 for ANOVAs and t-tests.
13.2.1. – Holocaust Survivor Gender
13.2.1.1 – Influence on survivor and descendant psychological health.
It was hypothesised that female survivors would score lower on positive variables and
higher on negative variables (DH1). A glance at Table 13.1 shows there is only one
statistically significant difference between the male and female survivors in the current
study sample with male survivors having a statistically significantly stronger belief in
world meaningfulness than female survivors.
The flow-on hypothesis argues that the descendants of the most affected ancestors
will be the most affected themselves (DH10). Given that it is hypothesised that female
survivors will be worse off than male survivors, it follows that it is also hypothesised that
children with only one survivor parent will fare less well if their survivor parent is their
mother as opposed to their father. However, it should also be noted that the one study that
directly examined this issue (Schleuderer, 1990) found that children of survivor fathers
scored less favourably than children of survivor mothers (as discussed in Chapters Nine and
Ten).
As can be seen in Table 13.1, once analyses controlling for confounds between
survivor parent gender and delay between 1945 and birth of children of survivors have been
conducted, there are no statistically significant differences between children of survivor
mothers and children of survivor fathers on psychological impact variables. However, it is
interesting to note that the pattern of results is consistent with having a survivor father
being more of a negative influence than having a survivor mother. This is consistent with
the findings of Schleuderer (1990), but inconsistent with the hypothesis that the children of
© Janine Lurie-Beck 2007
243
the most affected survivors should be the most affected themselves, which would suggest
children of survivor mothers should evidence higher symptom levels.
Table 13.1.
Influence of survivor gender on survivor and children of survivor scores on impact and influential process variables
Female
Survivors
(n = 11)
Male
Survivors
(n = 13)
Significance
Test Results
η2
Children
of
Survivor
Mother
only
(n = 5)
Children
of
Survivor
Father
only
(n = 12)
Significance Test Results
η2
4.00
(6.24)
5.54
(6.05)
t (21) = 0.60,
p = 0.56
0.017
1.20
(0.84)
5.83
(5.20)
0.201
ANCOVA
= 0.148
5.80
(5.69)
3.77
(2.64)
10.91
(4.65)
37.50
(9.86)
11.13
(4.12)
56.10
(26.30)
6.38
(6.98)
2.76
(2.92)
10.77
(4.13)
38.77
(7.41)
12.54
(3.64)
56.08
(30.40)
t (21) = 0.22,
p = 0.83
t (22) = 0.88,
p = 0.39
t (22) = 0.08,
p = 0.94
t (19) = 0.34,
p = 0.74
t (19) = 0.82,
p = 0.42
t (20) = 0.00,
p = 0.99
0.002
4.70
(6.02)
7.67
(9.13)
t (12.28) = 2.99, p < 0.05,
ANCOVA with delay as
covariate (F (1,14) = 2.43,
p = 0.141
t (15) = 0.66, p = 0.52
11.00
(5.29)
42.90
(11.76)
12.00
(5.66)
13.67
(3.94)
38.17
(10.30)
18.75
(7.88)
43.20
(7.60)
101.70
(15.36)
31.00
(7.28)
28.27
(7.55)
41.46
(9.62)
101.54
(16.88)
29.23
(6.80)
35.31
(9.40)
t (21) = 0.47,
p = 0.64
t (21) = 0.02,
p = 0.98
t (21) = 0.62,
p = 0.55
t (22) = 1.99,
p = 0.059
0.010
38.20
(9.88)
87.80
(21.29)
32.60
(5.94)
32.60
(4.62)
45.50
(10.67)
97.92
(16.63)
36.08
(4.10)
36.42
(6.64)
Impact Variables
DASS Anxiety
DASS Depression
IES-R Total Score
PTV
AAS Positive
Dimensions
AAS Negative
Dimensions
PTGI Total Score
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
0.034
0.000
0.006
0.034
0.000
0.000
0.017
0.153
0.028
Not applicable
t (15) = 1.15, p = 0.27
0.082
t (15) = 0.83, p = 0.42
0.044
t (15) = 1.73, p = 0.11
0.166
Not applicable
t (15) = 1.31, p = 0.21
0.103
t (15) = 1.06, p = 0.31
0.069
t (15) = 1.40, p = 0.18
0.116
t (15) = 1.16, p = 0.26
0.083
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale
13.2.1.2. – Influence on children of survivors’ perception of their parents/family
environment.
Further to the hypothesis about survivor gender is its hypothesised relationship with the
perceptions that the children have of their survivor parents. To be consistent with the flow
on hypothesis, it would be predicted that survivor mothers would be rated less favourably
than survivor fathers. Table 13.2 presents the ratings given by children whose parents are
both survivors of their survivor mother and their survivor father. As can be seen, there are
no statistically significant differences in ratings of survivor mothers and survivor fathers on
any of the dimensions measured. However it is interesting to note that survivor fathers are
rated less favourably on all three parent-child attachment dimensions, but more favourably
on their perceived level of support for their childrens’ independence.
© Janine Lurie-Beck 2007
244
Table 13.2.
Mean differences in ratings of survivor mothers versus survivor fathers among children with two survivor parents (n = 51)
on parent-child attachment dimensions and parental facilitation of independence
PCS - Warm
PCS - Cold
PCS - Ambivalent
PAQ – Fostering of Autonomy
HCQ - Frequent and Willing communication about the
Holocaust
HCQ - Guilt Inducing Communication
HCQ - Indirect Communication
HCQ - Affective Communication
Perception of Survivor
Mother
8.22 (5.10)
3.12 (4.06)
5.37 (4.79)
43.52 (14.18)
9.23 (2.93)
Perception of Survivor
Father
8.20 (5.13)
4.06 (4.31)
5.63 (4.48)
44.13 (14.92)
8.44 (3.29)
Significance Test
Results
t (48) = 0.03, p = 0.98
t (48) = 1.71, p = 0.10
t (48) = 0.35, p = 0.73
t (47) = 0.34, p = 0.74
t (47) = 1.36, p = 0.18
0.000
0.057
0.002
0.002
0.038
3.35 (1.76)
1.79 (1.04)
2.65 (1.02)
3.42 (1.74)
1.57 (0.88)
2.41 (0.98)
t (47) = 0.25, p = 0.81
t (46) = 1.30, p = 0.20
t (43) = 1.13, p = 0.26
0.001
0.035
0.029
η2
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire
Table 13.3 presents children of survivors’ ratings on family environment variables that are
not parent specific. In this case, it was necessary to compare the perceptions of children
who have a survivor mother only to those held by children who have a survivor father only.
There are no statistically significant differences.
Table 13.3.
Children with a survivor mother versus a survivor father only perceptions of family environment variables
FES – Cohesion
FES – Expressiveness
HCQ – Non-Verbal Presence of the
Holocaust
Children of Survivor Mother only
(n = 6)
49.17 (19.13)
38.33 (20.70)
8.50 (2.95)
Children of Survivor Father only
(n = 12)
41.92 (21.70)
42.17 (19.27)
9.08 (2.57)
Significance Test
Results
t (16) = 0.69, p = 0.50
t (16) = 0.39, p = 0.70
t (15) = 0.43, p = 0.67
η2
0.029
0.009
0.012
Notes. HCQ = Holocaust Communication Questionnaire, FES = Family Environment Scale
13.2.2. – Holocaust Survivor Age during the Holocaust
13.2.2.1. – Influence on survivor and descendant psychological health.
Survivor age during the Holocaust was operationalised by calculating the age of each
Holocaust survivor participant in 1945. It was hypothesised that negative effects would
increase with age (DH2). Correlations were conducted to assess the impact of age on
survivor scores on psychological impact and influential psychological process variables and
these are displayed in Table 13.4.
© Janine Lurie-Beck 2007
245
Table 13.4.
Correlations between survivor age in 1945 and impact and influential process variables among survivors and children of
survivors
Impact Variables
DASS Anxiety
DASS Depression
IES-R Total Score
PTV
AAS Positive Dimensions
AAS Negative Dimensions
PTGI Total Score
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
Holocaust Survivors
(n = 26)
Age in 1945
r
r2
Children of Survivors
Survivor Father’s age in 1945 (n = 62)
r
r2
-0.05
-0.19
0.45 *
-0.05
-0.06
-0.23
0.31
0.003
0.037
0.205
0.003
0.003
0.054
0.093
-0.004
0.265 *
0.000
0.070
-0.199
-0.024
-0.078
0.040
0.001
0.006
0.15
0.08
-0.21
-0.30
0.023
0.006
0.042
0.091
-0.037
-0.062
-0.157
-0.162
0.001
0.004
0.025
0.026
Survivor Mother’s age in 1945 (n = 55)
r
r2
Not applicable
Not applicable
0.031
0.075
0.001
0.006
-0.069
-0.033
0.058
0.005
0.001
0.003
-0.141
-0.220
-0.019
0.143
0.020
0.048
0.000
0.020
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale.
* p < 0.05.
There is a statistically significant positive relationship between PTSD symptoms
and survivor age during the Holocaust; however none of the other psychological impact or
influential psychological process variables correlate statistically significantly with survivor
age. The direction of this correlation is consistent with three of the four studies included in
the meta-analytic review of the literature to calculate correlations between survivor age and
PTSD symptoms (see Chapter Eight, Section 8.6).
Among the children of survivors, there is one statistically significant correlation
with survivor fathers’ age during the Holocaust with a positive relationship being found
between children of survivors’ depression scores and their survivor fathers’ age. As to
whether this result is in keeping with the meta-analytic review, it is consistent with the
results of Budick (1985) who found that depression scores among children increased with
the age of survivor parents, but is inconsistent with the findings of Eskin (1996) who found
a relationship in the opposing direction (see Chapter Nine, Section 9.2.6 for more
discussion of these studies). Overall, the correlations between survivor father age and
children of survivors’ scores are stronger than those with survivor mother age for the
current study sample.
13.2.2.2. – Influence on children of survivors’ perceptions of their parents/family
environment.
Table 13.5 presents a comparison of the strength of the relationships between survivor
fathers’ and survivor mothers’ age during the Holocaust and the perceptions held by their
© Janine Lurie-Beck 2007
246
children of each parent individually and the family environment. While there are no
statistically significant correlations, it is noteworthy that again it is the correlations with
paternal age that are stronger than for maternal age.
Table 13.5.
Correlations between Holocaust survivor parent age and children of survivors’ ratings of survivor parents on family
interaction variables
Survivor father’s age in 1945
(n = 62)
Parent specific family interaction variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Frequent and willing communication about the Holocaust
HCQ – Affective communication about the Holocaust
HCQ – Guilt-inducing communication about the Holocaust
HCQ – Indirect communication about the Holocaust
Non-Parent specific family interaction variables
HCQ – Non-verbal presence of the Holocaust
FES – Cohesion
FES – Expressiveness
Survivor mother’s age in 1945
(n = 55)
r
r2
r
r2
-0.118
0.178
0.013
-0.042
-0.201
0.066
-0.095
0.068
0.014
0.032
0.000
0.002
0.040
0.004
0.009
0.005
-0.095
0.012
0.057
-0.053
0.041
0.040
-0.016
0.146
0.009
0.000
0.003
0.003
0.002
0.002
0.000
0.021
0.251
-0.017
-0.043
0.063
0.000
0.002
0.098
-0.020
-0.130
0.010
0.000
0.017
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire, FES = Family Environment Scale
13.2.3. – Nature of Holocaust Experiences
13.2.3.1. – Influence on survivor and descendant psychological health.
The nature of Holocaust experiences is operationalised by camp internment, non-camp
experiences (mostly in hiding) and escape prior to 1945. It was hypothesised that camp
internment would lead to higher symptom levels than non-camp experiences such as hiding
or escaping (DH3). Due to the statistically significant relationship between the nature of
Holocaust experience and age during the Holocaust, ANCOVAs with age as a covariate
were conducted with the variables that were also statistically significantly correlated with
age. These ANCOVAs allow the reader to determine whether the statistically significant
differences between Holocaust experience groups in these cases are merely reflecting the
impact of age rather than a true difference between experience groups. This is the case
only for the IES-R total score.
As can be seen in Table 13.6, the statistically significant ANOVA result for
experience type is no longer statistically significant when conducted as an ANCOVA with
age in 1945 as a covariate. However, differences between survivor experience groups on
the Adult Attachment Scale do reach significance at the 0.10 level. Specifically, camp
survivors score statistically significantly lower on positive attachment dimensions than
© Janine Lurie-Beck 2007
247
survivors who escaped prior to 1945 and survivors who were in hiding or had other noncamp experiences score statistically significantly higher on negative attachment
dimensions/attachment anxiety than survivors who escaped prior to 1945. Within the
influential psychological process variables, three of the four variables differ statistically
significantly by survivor experience group. Specifically, camp survivors report statistically
significantly higher usage of maladaptive coping strategies than those who escaped prior to
1945 and cite statistically significantly weaker belief in world benevolence and
meaningfulness than survivors who escaped in hiding or in some other non-camp way. The
sizes of the η 2 cited in Table 13.6 are notable with many suggesting that a quarter or more
of the variation in scores can be explained by the type of experiences survivors had during
the Holocaust. While a small number of studies, cited in the meta-analytic review of the
literature, found that survivors who were hiding had/have higher levels of symptoms, the
findings of the current study are consistent with the majority of studies that have examined
the influence of the nature of Holocaust experiences on survivors’ mental health and
functioning (refer back to Chapter Eight, Section 8.2 for further details of the studies
reviewed).
Table 13.6.
Holocaust survivor experience group scores on impact and influential process variables
Impact Variables
DASS Anxiety
DASS Depression
IES-R Total Score
PTV
AAS Positive Dimensions
AAS Negative Dimensions
PTGI Total Score
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
Escape prior to 1945 a
(n = 6)
Camp b
(n = 10)
Hiding/ Other c
(n = 8)
Significance Test Results
η2
3.50 (5.50)
6.50 (8.17)
2.08 (2.10)
5.22 (7.93)
5.78 (6.24)
5.05 (2.90)
5.50 (4.34)
6.25 (5.75)
1.79 (1.81)
0.019
0.002
0.324
0.159
10.17 (4.17)
44.33 (8.73) b
9.00 (2.83) c
46.33 (27.38)
12.60 (4.35)
33.57 (4.65) a
11.57 (1.90)
68.11 (15.52)
9.13 (3.90)
37.88 (8.10)
14.63 (4.07) a
49.00 (37.62)
F (2,20) = 0.20, p = 0.82
F (2,20) = 0.02, p = 0.98
F (2,21) = 5.03, p < 0.05 ANOVA,
F (2,20) = 1.89, p = 0.177
ANCOVA with age in 1945 as
covariate,
F (2,21) = 1.65, p = 0.22
F (2,18) = 3.49, p = 0.052
F (2,18) = 5.60, p < 0.05
χ 2 Kruskal Wallis (2) = 1.91, p =
0.136
0.280
0.383
0.136
0.39
36.17 (6.43) b
107.50 (11.07)
31.17 (7.57)
31.83 (4.40)
46.78 (9.98) a
96.89 (18.52)
26.40 (7.29) c
27.30 (10.64) c
41.63 (5.78)
102.50 (15.92)
33.75 (3.54) b
38.25 (6.30) b
F (2,20) = 3.31, p = 0.057
F (2,20) = 0.81, p = 0.46
F (2,21) = 3.08, p = 0.067
χ 2 Kruskal Wallis (2) = 7.15, p <
0.249
0.075
0.227
0.277
0.05
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale. Lettered superscripts indicate which groups differ statistically significantly according to Tukey post-hoc analyses.
© Janine Lurie-Beck 2007
248
The flow on effects of the nature of survivors’ experience during the Holocaust on
their offspring have also been assessed. To be consistent with DH10 that the children of
the most affected survivors would be the most affected themselves, it is hypothesised that
children of camp survivors will fare less well than children of non-camp survivors.
Children of survivors are stratified by the nature of their survivor mother and survivor
fathers’ Holocaust experiences.
As can be seen in Table 13.7, children of survivor fathers who were in camps have
statistically significantly weaker belief in world benevolence than children of survivor
fathers who were in hiding or had some other non-camp experiences. In addition, children
of survivor fathers who were in hiding report the highest usage of adaptive coping
strategies, followed by children of camp survivor fathers and children of fathers who
escaped prior to 1945. The result pertaining to the world benevolence assumption is
consistent with the flow-on hypothesis. The findings of the meta-analytic review on this
issue were somewhat ambiguous, but did suggest a trend towards the children of non-camp
survivors evidencing higher symptomatology. Certainly a perusal of the mean scores in
Table 13.7 do not suggest a blanket trend towards the children of camp survivors
evidencing higher symptom levels.
However, the pattern of results do suggest that the
nature of survivor parents’ experiences during the Holocaust have some role to play.
© Janine Lurie-Beck 2007
249
Table 13.7.
Children of survivor scores on impact and influential process variables by survivor parent experience groups
Survivor Father’s Holocaust Experience
Escape prior to 1945
Camp
(n = 4) a
(n = 40) b
Hiding/Other
6.50 (6.45)
3.14 (3.41)
4.15 (4.39)
DASS Depression
14.75 (12.07)
7.33 (9.07)
PTV
12.00 (5.77)
11.28 (4.52)
Survivor Mother’s Holocaust Experience
η2
Escape prior to 1945
Camp
Hiding/ Other
Significance Test
(n = 5) a
(n = 35) b
(n = 11) c
Results
F (2,53) = 1.47,
0.054
3.80 (4.97)
3.10 (3.57)
2.27 (2.15)
F (2,48) = 0.39,
0.016
6.62 (9.07)
p = 0.24
F (2,53) = 1.25,
0.045
6.80 (5.81)
8.00 (10.39)
5.91 (6.43)
p = 0.68
F (2,48) = 0.22,
0.009
9.04 (5.08)
p = 0.30
F (2,53) = 1.43,
0.044
12.00 (4.85)
10.09 (4.51)
9.09 (5.55)
p = 0.80
F (2,48) = 0.64,
0.026
40.85 (10.64)
p = 0.25
F (2,53) = 0.72,
40.82 (10.61)
p = 0.53
F (2,48) = 0.09,
0.004
15.31 (6.24)
p = 0.49
F (2,53) = 1.83,
12.55 (4.25)
p = 0.91
F (2,48) = 0.39,
0.016
(n = 13)
c
Significance Test
Results
η2
Impact Variables
DASS Anxiety
AAS Positive Dimensions
AAS Negative Dimensions
34.25 (5.68)
21.00 (8.25)
40.18 (9.72)
14.63 (6.24)
0.027
0.063
38.60 (8.56)
14.60 (4.93)
40.31 (9.65)
14.17 (6.11)
p = 0.17
Influential Psychological
p = 0.68
Processes
COPE Maladaptive
44.25 (8.96)
41.38 (7.71)
39.10 (6.99)
F (2,53) = 0.78,
80.25 (11.70) c
95.98 (17.06)
101.54 (11.91) a
F (2,54) = 2.79,
WAS – Benevolence
36.50 (3.11)
32.77 (7.45) c
38.62 (5.92) b
F (2,53) = 3.65,
WAS – Meaningfulness
31.75 (1.26)
33.44 (8.03)
32.23 (7.52)
0.029
40.60 (3.91)
39.89 (6.32)
38.30 (7.70)
F (2,48) = 0.31,
0.094
90.00 (21.71)
93.35 (17.86)
99.36 (12.18)
F (2,48) = 0.69,
0.121
35.00 (2.83)
33.15 (8.79)
35.55 (5.26)
0.007
32.20 (8.98)
31.18 (8.13)
33.82 (5.33)
p = 0.73
p = 0.47
COPE Adaptive
p < 0.05
Wallis (2) = 1.38,
0.028
p = 0.51
p = 0.070
χ 2 Kruskal
0.013
χ
2
Kruskal Wallis
(2) = 0.26, p = 0.88
F (2,46) = 0.49,
0.019
0.021
p = 0.62
p = 0.50
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment Scale, WAS = World Assumptions Scale. Lettered superscripts
indicate which groups differ statistically significantly according to Tukey post-hoc analyses.
© Janine Lurie-Beck 2007
250
Among children whose parents are both survivors there is the question of how the
experiences of both parents interplay. To address this issue, a comparison was made
between children of survivors who have two parents who survived in hiding, two parents
who survived the camps and children of parents with differing experiences (see Table
13.8).
The one statistically significant difference resulting from this analysis group is that
children whose parents have differing Holocaust experiences score the lowest on positive
attachment dimensions (comfort with being close to and depending on others), followed by
children of two camp survivors and children of two hiding survivors.
Table 13.8.
Children of survivors’ scores on impact and influential process variables by survivor parent experience mixture groups
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
Both hiding/other
(n = 5) a
Both camps (n
= 26) b
Mixture
(n = 12) c
Significance testing
η2
2.00 (2.56)
3.40 (3.85)
6.50 (6.56)
48.60 (6.43) c
12.20 (1.79)
3.13 (3.78)
8.46 (10.32)
10.52 (4.73)
40.35 (10.32)
13.96 (6.08)
3.08 (2.84)
7.92 (9.98)
10.96 (3.29)
36.92 (7.90) a
15.00 (6.25)
F (2,40) = 0.24, p = 0.79
F (2,40) = 0.57, p = 0.57
F (2,40) = 1.83, p = 0.17
F (2,40) = 2.74, p = 0.077
F (2,40) = 0.41, p = 0.67
0.012
0.028
0.084
0.121
0.020
35.87 (6.27)
105.40 (3.96)
39.77 (5.98)
92.54 (17.45)
41.00 (6.65)
98.42 (16.83)
F (2,40) = 1.22, p = 0.31
0.058
0.071
38.20 (4.09)
34.80 (4.32)
32.04 (8.93)
31.56 (8.18)
35.75 (6.54)
32.17 (8.36)
χ
2
Kruskal Wallis (2) = 2.85, p =
0.24
F (2,39) = 1.75, p = 0.19
F (2,39) = 0.35, p = 0.71
0.082
0.018
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions Scale. Lettered superscripts indicate which
groups differ statistically significantly according to Tukey post-hoc analyses.
13.2.3.2. – Influence on children of survivors’ perception of their parents/family
environment.
To be consistent with the flow-on hypothesis (DH10), it is hypothesised that survivors who
were in camps would be rated less favourably by their children than survivors with noncamp experiences in terms of parent specific variables and general family
interaction/environment variables. As can be seen in Table 13.9, none of the parent specific
family interaction variables differ statistically significantly by survivor parent experience
category; however, the experiences of both survivor mother and father are statistically
significantly related to their children’s perceptions of a non-verbal presence of the
Holocaust when they were growing up. Specifically, mother and father camp experience is
related to elevated scores on this variable, as opposed to mother and father non-camp
experiences.
© Janine Lurie-Beck 2007
251
Table 13.9.
Children of survivors’ scores on family interaction variables by Holocaust experience of survivor parents
Survivor Father’s Holocaust Experience
Escape prior to
Camp
1945
(n = 40) b
Hiding/Other
(n = 13)
Significance Test Results
c
Survivor Mother’s Holocaust Experience
η2
(n = 4) a
Escape prior
Camp
to 1945
(n = 35) b
Hiding/ Other
(n = 11)
Significance Test Result
η2
c
(n = 5) a
Parent Specific Family
Interaction Variables
PCS – Warmth
8.75 (5.19)
9.10 (4.93)
7.58 (5.73)
F (2,51) = 0.32, p = 0.73
0.015
8.20 (3.27)
8.03 (5.52)
9.91 (5.87)
F (2,47) = 0.50, p = 0.61
0.021
PCS – Coldness
2.75 (2.99)
3.33 (4.12)
4.75 (4.45)
F (2,51) = 0.71, p = 0.50
0.024
2.40 (3.05)
3.32 (4.38)
1.82 (3.25)
F (2,47) = 0.61, p = 0.55
0.025
PCS – Ambivalence
6.50 (5.20)
4.54 (3.77)
6.58 (5.81)
χ
0.045
5.40 (3.44)
5.65 (5.13)
3.91 (5.24)
F (2,47) = 0.50, p = 0.61
0.021
45.00 (22.41)
46.21 (14.98)
44.50 (14.43)
1.02, p = 0.60
F (2,50) = 0.03, p = 0.97
0.002
47.40 (9.91)
41.76 (14.18)
46.00 (18.26)
F (2,47) = 0.55, p = 0.58
0.023
9.25 (1.71)
8.95 (3.45)
7.42 (2.81)
F (2,51) = 1.23, p = 0.30
0.041
8.20 (3.56)
9.18 (2.70)
9.55 (3.62)
F (2,42) = 1.68, p = 0.20
0.015
3.00 (1.41)
3.13 (1.65)
2.83 (1.27)
F (2,50) = 0.21, p = 0.82
0.007
3.40 (1.95)
3.59 (2.06)
3.00 (1.41)
F (2,47) = 0.39, p = 0.68
0.016
1.50 (1.00)
1.63 (0.85)
1.09 (0.30)
χ 2 Kruskal Wallis (2) =
0.075
1.20 (0.45)
1.82 (0.97)
1.55 (1.29)
F (2,47) = 0.99, p = 0.38
0.040
2.00 (0.82)
2.38 (0.93)
3.00 (0.87)
4.08, p = 0.13
F (2,46) = 2.13, p = 0.13
0.087
2.20 (1.10)
2.82 (1.00)
2.11 (0.78)
F (2,41) = 2.30, p = 0.11
0.101
FES – Cohesion
25.25 (19.36)
43.33 (22.33)
43.83 (22.56)
F (2,52) = 1.20, p = 0.31
0.045
47.20 (27.54)
39.94 (22.02)
45.55 (23.26)
F (2,46) = 0.40, p = 0.68
0.017
FES – Expressiveness
34.25 (19.14)
37.10 (19.04)
35.86 (17.20)
F (2,52) = 0.03, p = 0.98
0.002
38.20 (18.61)
34.00 (18.57)
32.57 (17.14)
F (2,47) = 0.17, p = 0.85
0.007
8.25 (0.96)
10.76 (2.88)
9.00 (3.05)
F (2,52) = 2.76, p = 0.073
0.096
10.00 (1.00)
11.01 (2.81)
8.20 (3.39)
F (2,46) = 3.86, p < 0.05
0.144
PAQ – Fostering of Autonomy
HCQ – Frequent and willing
2
Kruskal Wallis (2) =
communication about the
Holocaust
HCQ – Guilt-inducing
communication about the
Holocaust
HCQ – Indirect communication
about the Holocaust
HCQ – Affective communication
about the Holocaust
Non Parent Specific Family
Interaction Variables
HCQ – Non-verbal presence of
c
b
the Holocaust
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust Communication Questionnaire, FES = Family Environment Scale.
Lettered superscripts indicate which groups differ statistically significantly according to Tukey post-hoc analyses
© Janine Lurie-Beck 2007
252
Again, a comparison is made between the perceptions of children with two survivor
parents who had similar Holocaust experiences and children of two survivor parents with
differing Holocaust experiences. There are numerous statistically significant differences.
Maternal warmth is rated higher among children with two parents who survived in hiding.
Maternal indirect and frequent and willing communication about the Holocaust is also rated
higher among children with two parents who were in hiding. By contrast, paternal affective
communication is perceived to be higher among children whose survivor parents’
Holocaust experiences differed, while frequent and willing communication about the
Holocaust is rated higher among children of two camp survivors. Finally, non-verbal
presence of the Holocaust is also perceived to be higher in two camp survivor families.
Table 13.10.
Children of survivors’ scores on family interaction variables by survivor parent experience mixture groups
Maternal Variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication
about the Holocaust
HCQ – Indirect communication
about the Holocaust
HCQ – Guilt-inducing
communication about the
Holocaust
HCQ – Frequent and willing
communication about the
Holocaust
Paternal Variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication
about the Holocaust
HCQ – Indirect communication
about the Holocaust
HCQ – Guilt-inducing
communication about the
Holocaust
HCQ – Frequent and willing
communication about the
Holocaust
Non parent specific family
interaction variables
HCQ – Non-verbal presence of the
Holocaust
FES – Cohesion
FES – Expressiveness
Both
hiding/other
(n = 5) a
Both camps
(n = 26) b
Mixture
(n = 12) c
13.80 (1.92)
7.46 (5.69)
8.18 (4.83)
0.80 (1.30)
0.40 (0.55)
56.40 (8.65)
1.75 (0.50)
3.88 (4.76)
5.69 (4.91)
41.62 (15.35)
2.78 (0.98)
1.82 (3.46)
5.18 (4.60)
41.18 (14.08)
2.60 (1.17)
2.20 (1.79)
1.92 (0.98)
1.18 (0.60)
2.40 (0.55)
3.85 (2.19)
3.27 (1.68)
12.00 (2.83)
9.52 (2.31)
8.80 (5.76)
6.20 (5.22)
6.60 (5.37)
47.40 (18.47)
2.33 (0.58)
Significance testing
Kruskal Wallis
χ 2 (2) = 5.93, p = 0.052
η2
0.138
F (2,39) = 1.69, p = 0.20
F (2,39) = 1.75, p = 0.19
F (2,39) = 2.34, p = 0.11
F (2,36) = 1.83, p = 0.18
0.080
0.127
0.107
0.092
χ 2 (2) = 5.85, p = 0.054
2
Kruskal Wallis χ (2) = 1.81, p = 0.40
0.117
7.64 (3.50)
F (2,38) = 4.58, p < 0.05
0.194
8.12 (5.10)
3.88 (4.54)
4.92 (4.35)
42.88 (16.64)
2.22 (0.86)
8.09 (5.52)
3.18 (3.82)
6.00 (5.12)
43.90 (11.08)
3.20 (1.03)
F (2,38) = 0.04, p = 0.96
F (2,38) = 0.81, p = 0.45
F (2,38) = 0.39, p = 0.68
F (2,37) = 0.17, p = 0.84
F (2,33) = 4.22, p < 0.05
0.002
0.041
0.020
0.009
0.204
1.00 (0.00)
1.67 (0.87)
1.27 (0.65)
2.40 (0.55)
3.17 (1.66)
3.73 (1.95)
7.00 (2.45)
9.16 (3.44)
6.64 (2.94)
F (2,38) = 2.76, p = 0.076
0.127
7.80 (3.63)
11.48 (2.66)
9.70 (2.71)
F (2,38) = 4.33, p < 0.05
0.186
47.20 (21.94)
34.46 (19.35)
38.69 (22.32)
33.50 (18.53)
44.91 (25.55)
34.73 (15.86)
F (2,39) = 0.46, p = 0.63
F (2,39) = 0.02, p = 0.98
0.023
0.001
Kruskal Wallis
Kruskal Wallis
Kruskal Wallis
χ 2 (2) = 3.82, p = 0.15
χ 2 (2) = 1.38, p = 0.50
0.061
0.094
0.057
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire, FES = Family Environment Scale. Lettered superscripts indicate which groups differ
statistically significantly according to Tukey post-hoc analyses.
© Janine Lurie-Beck 2007
253
13.2.4. – Loss of Family
The assessment of the impact of loss of family was possible from two angles. Firstly, it
was determined whether each survivor participant believed themselves to be the sole
survivor of their family. However, it is also true for many that although other family
members survived, the survivors found themselves separated from other family members
for at least part of their Holocaust experiences. So the two issues here are whether the
survivor was alone after the war and whether the survivor was ever alone during their
Holocaust traumas. It was hypothesised that greater negative effects would be associated
with greater losses of family (DH4).
13.2.4.1. – Influence on survivor and descendant psychological health.
Two sets of t-tests have been conducted to assess these two issues within the survivor
sample (see Table 13.11). The first set of t-tests compares survivors who are/were the sole
surviving member of their family at the end of the war and survivors for whom at least
some family members survived. There is one statistically significant difference between
sole-survivors and non-sole-survivors with non-sole survivors scoring statistically
significantly higher on depression than sole-survivors. This is the direct opposite of what
was predicted. However, it is interesting to note that in all other variables (apart from
anxiety which has a similar result to depression), it is the sole-survivors who score less
favourably than the non-sole-survivors albeit not statistically significantly. Indeed the
pattern here for the psychopathological variables suggests the presence of PTSD symptoms
among sole survivors, but PTSD symptoms with signs of co-morbid depression and anxiety
among the survivors with surviving family members. While it is intuitive to predict that
sole survivors would score higher on measures of psychopathology, this result is in keeping
with the results of Hafner (1968, discussed in Chapter Eight section 8.4) who found higher
incidence rates of psychopathological symptoms among survivors with surviving family
members. He hypothesised that interacting with surviving family members exacerbated
their symptoms. Unfortunately, a second study to consider sole-survivor status failed to
report descriptive statistics when a non-statistically significant result was obtained, so no
trends can be noted (Silow, 1993). The reader is invited to refer back to Chapter Eight,
Section 8.4 for further information on these studies.
The second set of t-tests which compare survivors who were always with at least
one family member during the war to survivors who for at least part of the war were
separated from their family yields one statistically significant difference. Survivors who
© Janine Lurie-Beck 2007
254
were alone at some stage during the war report much higher post-traumatic growth than
survivors who were always accompanied by family members. Again the one study located
for the review of the literature in Chapter Eight, Section 8.4, which considered time spent
alone by survivors (Cordell, 1980), obtained no statistically significant result but reported
no descriptive statistics.
The finding of the current study of higher post-traumatic growth among survivors
who had spent some time alone during the Holocaust is intuitively understandable. Being
forced to face their situation on their own, these survivors may have initially doubted their
ability to cope and survive but in hindsight would have no choice but to acknowledge that
they had been able to. Such an acknowledgement would have included a recognition of
personal strength, the realisation of which would no doubt have remained with them after
the Holocaust.
Table 13.11.
Holocaust Survivor scores on impact and influential process variables by loss of family variables
Sole surviving member of family after the Holocaust
Spent time without any family
members during the Holocaust
Yes
No
Significance
(n = 10)
(n = 9)
Test Results
No
(n = 20)
Significance
Test Results
η2
Impact Variables
DASS Anxiety
Yes
(n = 3)
0.33 (0.58)
5.55 (6.20)
0.088
5.10 (5.74)
6.22 (7.29)
DASS Depression
0.67 (0.58)
6.95 (6.39)
0.117
5.70 (6.06)
8.67 (7.02)
IES-R Total Score
3.60 (2.70)
3.17 (2.86)
0.003
2.75 (2.68)
3.69 (3.30)
10.50
(0.71)
30.00
(4.24)
12.50
(3.54)
56.00
(8.49)
10.86
(4.47)
39.16
(8.08)
11.95
(3.91)
56.10
(29.31)
t (21) = 1.43,
p = 0.17
t (20.55) = 4.29,
p < 0.001
t (22) = 0.24,
p = 0.81
t (22) = 0.11,
p = 0.91
t (19) = 1.55,
p = 0.14
t (19) = 0.19,
p = 0.85
t (20) = 0.01,
p = 0.99
0.001
9.46 (3.82)
0.113
39.67 (5.87)
0.002
13.11 (3.95)
0.000
71.20 (21.46)
40.33
(4.62)
96.67
(14.19)
29.33
(6.43)
25.00
(5.00)
42.50
(9.16)
102.35
(16.32)
30.14
(7.14)
33.10
(9.23)
t (21) = 0.40,
p = 0.70
t (21) = 0.57,
p = 0.58
t (22) = 0.19,
p = 0.86
t (22) = 1.47,
p = 0.16
0.007
41.60 (6.90)
0.015
103.50 (12.27)
0.002
31.82 (5.78)
0.089
33.36 (10.49)
PTV
AAS Positive Dimensions
AAS Negative Dimensions
PTGI Total Score
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
η2
0.008
12.50
(5.09)
38.56
(10.57)
10.67
(3.91)
38.11
(29.29)
t (17) = 0.38,
p = 0.71
t (17) = 0.99,
p = 0.34
t (18) = 70,
p = 0.49
t (19) = 1.57,
p = 0.13
t (16) = 0.28,
p = 0.79
t (16) = 1.32,
p = 0.21
t (17) = 2.83,
p < 0.05
43.78
(11.91)
104.44
(18.47)
29.00
(8.49)
32.33
(8.53)
t (17) = 0.49,
p = 0.63
t (17) = 0.13,
p = 0.90
t (18) = 0.88,
p = 0.39
t (18) = 0.24,
p = 0.82
0.014
0.054
0.027
0.115
0.005
0.098
0.320
0.001
0.041
0.003
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale
The reverberation of survivors’ loss of family during the Holocaust can also be
examined. Children of survivors have been sub-divided based on whether their survivor
parent was the sole survivor of their family. This issue was addressed separately for
© Janine Lurie-Beck 2007
255
survivor fathers and survivor mothers. As can be seen in Table 13.12, there are two
statistically significant results for this issue and both are in the direction that would be
predicted. Specifically children of sole-survivor mothers report higher depression scores
and lower positive attachment dimensions scores than children whose mothers were not
sole survivors. These results are intuitively predictable and consistent with the two studies
reported in the review of the literature in Chapter Nine, Section 9.2.4 (Gertler, 1986;
Schwarz, 1986). The reader should note that there is no potential for confound with the
sole-survivorship status of the other parent, as there is only one child of two sole survivors
in the sample and this participant was removed from the analyses. Therefore all analyses
presented in Table 13.12 assume the other parent is a not a sole survivor.
Table 13.12.
Children of survivor’ scores on impact and influential process variables by survivor parents’ loss of family during the
Holocaust
Sole Survivor
(n = 12)
Impact Variables
DASS Anxiety
Survivor Father
Not Sole
Significance
Survivor
Test Results
(n = 38)
4.67 (3.65)
3.57 (4.17)
DASS
Depression
PTV
11.08 (13.06)
7.79 (8.27)
13.08 (3.78)
11.07 (4.54)
AAS Positive
Dimensions
AAS Negative
Dimensions
Influential
Psychological
Processes
COPE
Maladaptive
COPE Adaptive
36.83 (12.52)
39.53 (8.87)
18.25 (8.14)
14.97 (5.98)
42.00 (9.18)
41.34 (7.30)
97.25 (15.94)
95.09 (17.70)
WAS –
Benevolence
WAS –
Meaningfulness
31.83 (6.28)
34.03 (7.49)
33.33 (8.76)
32.78 (7.81)
η
2
Sole Survivor
(n = 3)
Survivor Mother
Not Sole
Survivor
(n = 46)
t (48) = 0.82,
p = 0.42
t (13.89) =
0.82, p = 0.42
t (48) = 1.39,
p = 0.17
t (48) = 0.83,
p = 0.41
t (48) = 1.51,
p = 0.14
0.014
4.00 (4.58)
2.79 (3.40)
0.022
23.00 (14.18)
6.55 (8.22)
0.039
6.33 (2.08)
10.17 (4.81)
0.014
30.33 (4.73)
40.86 (9.64)
0.046
16.33 (10.50)
13.83 (5.36)
t (48) = 0.26,
p = 0.80
t (48) = 0.38,
p = 0.71
t (47) = 0.91,
p = 0.37
t (47) = 0.21,
p = 0.84
0.001
42.67 (6.81)
39.31 (6.52)
0.003
80.33 (16.50)
94.22 (16.69)
0.017
36.00 (2.65)
33.25 (7.79)
0.001
29.33 (11.72)
31.84 (7.55)
Significance
Test Results
η2
t (47) = 0.59,
p = 0.56
t (47) = 3.23,
p < 0.01
t (47) = 1.37,
p = 0.18
t (47) = 1.86,
p = 0.069
t (47) = 0.74,
p = 0.46
0.007
t (47) = 0.86,
p = 0.39
t (47) = 1.40,
p = 0.17
t (45) = 0.60,
p = 0.55
t (45) = 0.54,
p = 0.59
0.016
0.181
0.038
0.069
0.012
0.040
0.008
0.006
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
13.2.4.2. – Influence on children of survivors’ perceptions of their parents/family
environment.
Whether the fact that a survivor was the sole survivor of his or her family after the
Holocaust has an impact on the perceptions their children have of their survivor parents is
examined in this sub-section. Given that it is hypothesised that sole survivors have been
more negatively affected by the Holocaust than non-sole survivors, one would expect
© Janine Lurie-Beck 2007
256
children of survivors to rate sole survivor parents less favourably than non-sole survivor
parents. While overall sole survivor status does seem to be related to less favourable
ratings (see Table 13.13), there are a few perplexing findings as well. In particular, it
seems that while sole survivor mothers are rated as more ambivalent than non-sole survivor
mothers, sole survivor fathers are rated as less ambivalent than non-sole survivor fathers.
There are no other studies, that were located, that considered the role of this variable among
children of survivors’ perceptions of family environment.
Table 13.13
Children of survivors’ scores on family interaction variables by sole-survivor status of survivor parents
Yes
(n = 12)
Survivor Father
No
Significance Test
(n = 38)
Results
Parent specific
family
interaction
Variables
PCS – Warmth
8.09 (5.26)
8.42 (4.76)
PCS – Coldness
3.27 (4.00)
3.42 (3.74)
PCS –
Ambivalence
PAQ – Fostering
of Autonomy
HCQ – Frequent
and willing
communication
about the
Holocaust
HCQ – Guiltinducing
communication
about the
Holocaust
HCQ – Indirect
communication
about the
Holocaust
HCQ – Affective
communication
about the
Holocaust
Non parent
specific family
interaction
variables
FES – Cohesion
3.36 (2.62)
5.84 (4.65)
45.80 (10.38)
45.24 (14.56)
9.55 (4.08)
8.89 (2.89)
4.00 (1.95)
FES –
Expressiveness
HCQ – Nonverbal presence
of the Holocaust
η
2
Yes
(n = 3)
Survivor Mother
No
Significance
(n = 46)
Test Results
t (47) = 0.20,
p = 0.84
t (47) = 0.11,
p = 0.91
t (29.89) = 2.27,
p < 0.05
t (46) = 0.11,
p = 0.91
t (47) = 0.60,
p = 0.55
0.001
6.33 (1.53)
8.61 (5.59)
0.000
3.67 (2.08)
2.83 (4.18)
0.057
10.00 (3.00)
4.83 (5.01)
0.000
29.67 (6.43)
44.17 (14.91)
0.008
11.00 (4.00)
9.16 (2.87)
2.87 (1.38)
t (47) = 2.18,
p < 0.05
0.092
3.67 (2.08)
2.00 (1.00)
1.37 (0.67)
t (12.75) = 1.97,
p = 0.071
0.113
2.50 (0.85)
2.50 (0.92)
t (44) = 0.00,
p = 1.00
44.33 (26.54)
39.74 (21.24)
40.58 (16.98)
33.75 (16.95)
11.91 (3.36)
9.99 (2.62)
t (48) = 0.62,
p = 0.54
t (48) = 1.22,
p = 0.23
t (47) = 2.01,
p = 0.050
ANCOVA with
number of
parents
F (1,46) = 5.89,
p < 0.05
η2
t (6.78) = 1.89,
p = 0.10
t (47) = 0.34,
p = 0.73
t (47) = 1.76,
p = 0.086
t (47) = 1.66,
p = 0.10
t (46) = 1.06,
p = 0.30
0.010
3.48 (1.92)
t (47) = 0.16,
p = 0.87
0.001
2.00 (1.00)
1.72 (1.03)
t (47) = 0.46,
p = 0.65
0.005
0.000
3.00 (1.41)
2.56 (0.99)
t (40) = 0.60,
p = 0.55
0.009
0.008
18.67 (24.01)
43.07 (21.88)
0.070
0.030
36.33 (9.71)
33.12 (17.98)
0.079
0.114
11.33 (1.15)
10.34 (3.10)
t (46) = 1.86,
p = 0.069
t (47) = 0.31,
p = 0.76
t (46) = 0.55,
p = 0.59
0.003
0.062
0.055
0.024
0.002
0.006
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire, FES = Family Environment Scale
© Janine Lurie-Beck 2007
257
13.2.5. – Holocaust Survivors’ Country of Origin
13.2.5.1. – Influence on survivor and descendant psychological health.
A survivor’s country of origin may have been indirectly influential in the nature of the postwar adjustment. This is because the period of persecution differed depending on the
country a survivor was from. Because experiences during the Holocaust depended on the
survivor’s country of origin, it was hypothesised that statistically significant differences in
psychological health would be noted when survivors are stratified by country of origin
(DH6). The planned method of analysis was to conduct ANOVAs with a survivor’s
country of origin as the independent variable. This was not possible due to very small
sample sizes for some countries. Therefore, survivors are grouped according to regions
created by the researcher to reflect similar wartime experiences. These groupings are:
Germany and Austria; Poland, Latvia and Lithuania; Belgium and the Netherlands; and
Hungary.
Table 13.14 reports the group means and results of significance testing examining
the role of survivors’ country of origin. As was the case for the nature of Holocaust
experiences, the effect sizes for a number of variables suggest that quite a sizeable
proportion of variance in scores are associated with country of origin (five of the η 2 reveal
at least a quarter of the variance is explainable), suggesting it is a variable of a fair degree
of import. With regard to the statistically significant differences, it appears that survivors
from Hungary have a heightened sense of vulnerability and a much weaker faith in world
meaningfulness than survivors from other countries. However in the area of attachment
anxiety, it is eastern European survivors who fare the worst. There were no studies located
for the meta-analysis that considered survivor country of origin so no comparison can be
made to the current study data.
© Janine Lurie-Beck 2007
258
Table 13.14.
Holocaust survivor scores on impact and influential process variables by country of origin
Impact Variables
DASS Anxiety
DASS Depression
IES-R Total Score
PTV
AAS Positive
Dimensions
AAS Negative
Dimensions
PTGI Total Score
Influential
Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS –
Meaningfulness
Germany,
Austria
(n = 10) a
Poland, Lithuania,
Ukraine
(n = 5) b
Netherlands,
Belgium
(n = 4) c
Hungary
(n = 4) d
Significance Test Results
5.80 (7.07)
7.00 (7.75)
3.75 (2.06)
1.00 (1.15)
7.30 (7.42)
3.59 (3.00)
11.56 (4.07)
40.40 (8.91)
5.60 (6.66)
2.67 (3.75)
9.75 (2.40)
32.75 (6.95)
6.00 (6.38)
1.28 (1.20)
6.63 (1.80) d
43.00 (5.89)
4.00 (4.08)
4.60 (1.54)
14.20 (5.26) c
32.33 (3.21)
2.67, p = 0.45
F (3,19) = 0.25, p = 0.86
F (3,20) = 1.21, p = 0.33
F (3,20) = 3.30, p < 0.05
F (3,17) = 2.09, p = 0.14
0.038
0.154
0.331
0.269
10.10 (2.81) b
16.50 (1.29) a
13.00 (5.35)
11.00 (1.00)
F (3,17) = 4.29, p < 0.05
0.431
47.56 (30.93)
63.60 (30.62)
60.33 (39.27)
61.40 (13.89)
F (3,18) = 0.45, p = 0.72
0.069
40.90 (10.44)
103.40 (19.44)
30.20 (6.36)
34.20 (6.03) d
40.80 (5.67)
102.00 (12.35)
30.40 (5.03)
32.40 (6.47) d
42.25 (7.80)
104.25 (14.73)
35.50 (4.04)
41.75 (4.99) d
47.25 (8.69)
94.00 (14.05)
25.00 (9.25)
19.80 (6.76) a b c
F (3,19) = 0.54, p = 0.66
F (3,19) = 0.35, p = 0.79
F (3,20) = 1.93, p = 0.16
F (3,20) = 10.40, p < 0.001
0.078
0.053
0.224
0.609
Kruskal Wallis
χ 2 (3) =
η2
0.120
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale. Lettered superscripts indicate which groups differ statistically significantly according to Tukey post-hoc analyses.
The focus now turns to whether a survivor’s country of origin can explain
differences within the children of survivor population. Certainly this appears to be the case
when it comes to strength of belief in world benevolence with children of Hungarian
survivor mothers and fathers reporting statistically significantly weaker beliefs than
children of survivors from other countries (see Tables 13.15 and 13.16). Children of
Hungarian survivor fathers also record the highest vulnerability levels. These findings for
the child of survivor generation mirror those for the survivor generation, with Hungarian
ancestry related to the most negative outcomes. Why survivors (and their descendants)
from Hungary would be particularly susceptible to higher levels of pathological symptoms
can perhaps be explained by the speed with which survivors (particularly Jews) were
moved into concentration camps. In most other countries, the process was more gradual
with a period of ghettoisation preceding camp internment. The transportation of Jews to
camps in Hungary was quite late in the war and was more hurried.
The gradual
acclimatisation to regimented living provided by a period of ghetto living has been
suggested as enabling the survivor to develop resilience which would be somewhat
protective in the camp environment (see Chapter Four, Section 4.7). If a more gradual
increase in persecution allowed survivors to get used to their living conditions and
treatment then the lack of this experienced by survivors in Hungary can certainly explain
© Janine Lurie-Beck 2007
259
the results found here. This is not to say that other factors may not play a role such as
cultural, ethnic differences in pre-existing resilience.
Table 13.15.
Children of survivors’ scores on impact and influential process variables by survivor mother country of origin
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
Germany,
Austria
(n = 6) a
Poland, Lithuania,
Ukraine, Latvia
(n = 29) b
Netherlands
(n = 4 ) c
Hungary
(n = 9) d
Significance Test Results
η2
3.17 (4.71)
5.50 (6.02)
9.00 (4.98)
38.33 (6.68)
14.33 (4.46)
3.16 (3.67)
8.02 (10.75)
9.50 (4.83)
41.16 (10.42)
14.21 (6.01)
1.75 (2.22)
5.75 (9.60)
6.88 (5.27)
45.75 (10.50)
13.25 (7.27)
1.89 (1.83)
7.22 (7.80)
12.67 (3.32)
38.56 (8.88)
13.22 (5.87)
F (3,44) = 0.45, p = 0.72
F (3,44) = 0.16, p = 0.93
F (3,44) = 1.77, p = 0.17
F (3,44) = 0.64, p = 0.60
F (3,44) = 0.09, p = 0.97
0.030
0.010
0.108
0.042
0.006
39.83 (5.95)
79.83 (16.09) b
36.83 (3.82) d
33.33 (8.50)
39.55 (6.29)
98.20 (17.04) a
34.64 (6.90) d
31.39 (8.12)
34.83 (6.26)
101.92 (11.59)
39.50 (2.38) d
34.75 (6.90)
39.89 (6.94)
85.89 (14.30)
27.11 (8.72) a b c
32.00 (6.98)
F (3,44) = 0.71, p = 0.55
F (3,44) = 3.27, p < 0.05
F (3,43) = 4.39, p < 0.01
F (3,43) = 0.27, p = 0.84
0.046
0.182
0.234
0.019
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale. Lettered superscripts indicate which groups differ statistically significantly
according to Tukey post-hoc analyses.
Table 13.16.
Children of survivors’ scores on impact and influential process variables by survivor father country of origin
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative
Dimensions
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
Germany,
Austria
(n = 11) a
Poland, Lithuania
(n = 30) b
Netherlands
(n = 3) c
Hungary,
Yugoslavia
(n = 10) d
Significance Test Results
η2
4.55 (4.63)
7.91 (8.78)
12.91 (3.78)
39.82 (7.21)
19.27 (7.35)
4.02 (4.12)
7.67 (10.64)
9.80 (4.97)
39.47 (10.98)
14.10 (6.30)
2.00 (2.65)
6.67 (11.55)
6.50 (6.38)
44.00 (12.12)
15.67 (6.66)
2.20 (2.49)
8.40 (7.38)
13.30 (3.40)
40.00 (9.42)
14.50 (5.80)
F (3,50) = 0.90, p = 0.45
F (3,50) = 0.03, p = 0.99
F (3,50) = 3.13, p < 0.05
F (3,50) = 0.18, p = 0.91
F (3,50) = 1.79, p = 0.16
0.051
0.002
0.158
0.011
0.097
41.91 (7.09)
94.18 (19.30)
35.36 (3.80) d
41.00 (7.37)
99.41 (16.36)
35.55 (6.29) d
35.78 (7.32)
98.89 (12.10)
40.00 (2.65) d
42.30 (10.04)
86.60 (13.68)
27.20 (8.95) a b c
F (3,50) = 0.58, p = 0.63
F (3,50) = 1.60, p = 0.20
0.033
0.088
0.249
34.27 (5.35)
32.52 (7.76)
37.67 (4.51)
32.00 (10.19)
Kruskal Wallis
χ 2 (3) =
10.26,
p < 0.05
F (3,49) = 0.55, p = 0.65
0.033
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale. Lettered superscripts indicate which groups differ statistically significantly
according to Tukey post-hoc analyses.
13.2.5.2. – Influence on children of survivors’ perception of their parents/family
environment.
The question of whether a survivor’s country of origin is related to their children’s
perception of them as parents is addressed in this section. Children of Dutch survivors are
statistically significantly less likely to note a non-verbal presence of the Holocaust in their
home.
However the third statistically significant result has Dutch fathers rated as
© Janine Lurie-Beck 2007
260
statistically significantly colder than survivor fathers from other regions. Whether this is a
result of Holocaust experiences or pre-existing cultural differences cannot be determined.
Table 13.17.
Children of survivors’ scores on family interaction variables by survivor mother country of origin
Parent specific family interaction
variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication
about the Holocaust
HCQ – Guilt-inducing
communication about the
Holocaust
HCQ – Indirect communication
about the Holocaust
HCQ – Frequent and willing
communication about the
Holocaust
Non parent specific family
interaction variables
FES – Cohesion
FES – Expressiveness
HCQ – Non-verbal presence of the
Holocaust
Germany,
Austria
(n = 6) a
Poland, Lithuania,
Ukraine, Latvia
(n = 29) b
Netherlands
(n = 4 ) c
Hungary
(n = 9) d
Significance Test Results
η2
9.00 (3.52)
2.00 (2.90)
4.33 (3.93)
49.00 (10.26)
2.00 (1.22)
8.17 (5.04)
2.86 (3.95)
5.24 (5.10)
41.93 (14.74)
2.69 (0.93)
11.75 (6.65)
3.25 (5.19)
2.75 (4.86)
51.25 (18.46)
2.00 (0.00)
11.00 (5.50)
1.89 (3.14)
4.00 (4.47)
48.11 (12.58)
3.14 (1.21)
F (3,44) = 1.09, p = 0.36
F (3,44) = 0.24, p = 0.87
F (3,44) = 0.41, p = 0.75
F (3,44) = 1.00, p = 0.40
F (3,38) = 1.73, p = 0.18
0.069
0.016
0.027
0.064
0.120
3.17 (1.83)
3.62 (2.04)
3.25 (1.89)
2.89 (1.36)
F (3,44) = 0.39, p = 0.76
0.026
1.50 (0.84)
1.86 (1.09)
1.50 (1.00)
1.78 (1.09)
F (3,44) = 0.29, p = 0.83
0.019
8.83 (3.76)
8.72 (2.95)
9.75 (4.11)
9.63 (3.02)
F (3,43) = 0.26, p = 0.86
0.018
45.67 (24.92)
37.50 (21.97)
9.50 (2.35)
42.89 (23.28)
34.08 (17.43)
11.22 (2.63) c
36.25 (25.20)
37.25 (16.36)
5.67 (3.06) b
41.56 (25.07)
37.78 (23.02)
8.89 (3.02)
F (3,43) = 0.13, p = 0.94
F (3,44) = 0.13, p = 0.94
F (3,43) = 5.06, p < 0.01
0.009
0.009
0.261
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire, FES = Family Environment Scale. Lettered superscripts indicate which groups differ
statistically significantly according to Tukey post-hoc analyses.
Table 13.18.
Children of survivors’ scores on impact and influential process variables by survivor father country of origin
Parent specific family interaction
variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication
about the Holocaust
HCQ – Guilt-inducing
communication about the
Holocaust
HCQ – Indirect communication
about the Holocaust
HCQ – Frequent and willing
communication about the
Holocaust
Non parent specific family
interaction variables
FES – Cohesion
FES – Expressiveness
HCQ – Non-verbal presence of
the Holocaust
Germany,
Austria
(n = 11) a
Poland, Lithuania
(n = 30) b
Netherlands
(n = 3) c
Hungary,
Yugoslavia
(n = 10) d
Significance Test Results
η2
10.09 (3.67)
1.55 (2.21) c
4.36 (3.78)
49.82 (14.88)
2.45 (0.93)
8.50 (5.62)
3.87 (4.46)
4.67 (4.10)
44.43 (15.50)
2.50 (0.99)
5.00 (5.00)
8.67 (1.15) a
10.00 (2.65)
41.50 (23.33)
2.00 (0.00)
9.89 (4.70)
3.11 (4.28)
5.11 (5.88)
49.00 (13.51)
2.56 (1.05)
F (3,49) = 0.96, p = 0.42
F (3,49) = 2.68, p = 0.057
F (3,49) = 1.47, p = 0.24
F (3,48) = 0.50, p = 0.68
F (3,44) = 0.19, p = 0.91
0.055
0.141
0.082
0.031
0.013
2.82 (1.17)
3.07 (1.56)
3.67 (2.08)
2.89 (1.69)
F (3,48) = 0.27, p = 0.85
0.017
1.27 (0.65)
1.72 (0.88)
1.00 (0.00)
1.11 (0.33)
χ 2 (3) =
0.126
Kruskal Wallis
9.09 (2.47)
8.27 (3.51)
7.00 (3.46)
8.89 (3.02)
6.28,
p = 0.099
F (3,49) = 0.43, p = 0.73
46.27 (22.28)
39.45 (18.05)
9.00 (1.67) c
43.57 (23.26)
36.38 (18.94)
11.05 (2.85) c
28.33 (24.01)
29.67 (7.51)
4.00 (1.41) a b d
37.50 (22.43)
34.90 (22.04)
10.20 (3.05) c
F (3,50) = 0.66, p = 0.58
F (3,50) = 0.24, p = 0.87
F (3,49) = 5.34, p < 0.01
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire, FES = Family Environment Scale. Lettered superscripts indicate which groups differ
statistically significantly according to Tukey post-hoc analyses.
© Janine Lurie-Beck 2007
261
0.026
0.038
0.014
0.246
The overall trend appears to be that both survivors and children of survivors from
the Netherlands score themselves more positively than those from other countries,
especially Hungary. However, it is difficult to disentangle the specifics of which survivor
countries of origin are the most predictive of negative outcomes for survivors and children
of survivors solely on this data. Larger sample sizes and more countries are required for a
more definitive assessment of this issue. What is clear, however, is that there is certainly an
argument for survivor country of origin being influential not only on the survivors
themselves but also their offspring. This does not necessarily relate to a linear relationship
with the number of years of persecution experienced in each country (which is open to
debate), but perhaps more to the cultural, economic, social, religious and political climates
of these countries before, during and after the war.
13.2.6. – Length of Time after 1945 Before Survivor Resettlement
What is of interest in this subsection is the potential negative influence of a drawn-out
emigration process for the survivors who wanted to emigrate from Europe after the war.
13.2.6.1. – Influence on survivor and descendant psychological health.
The amount of time that survivors spent “in limbo” in Europe immediately after the war
and before they arrived in their final resettlement location has been mentioned by numerous
authors as being an important factor in the post-Holocaust recovery of survivors. Much of
this time would have been spent in displaced persons camps or DP camps, which, as has
been discussed in numerous chapters already, was a potentially compounding traumatic
event (given the similarities in conditions between many DP camps and the concentration
camps the survivors had just been liberated from). It is thought that a longer delay in
resettlement, (which in many cases involves a longer duration of time spent in DP camps)
would have a more detrimental influence on survivors. The flow on effect of this waiting
time in Europe on the children of survivors is also examined in this section.
To investigate the influence of the amount of time survivors waited in Europe
before resettlement, correlations between the number of years spent in Europe between
1945 and the year survivors settled in their new home and the psychological impact and
influential psychological process variables have been calculated. However, as can be seen
in Table 13.19 there are no statistically significant relationships.
© Janine Lurie-Beck 2007
262
Table 13.19.
Correlations between survivor time in Europe before emigration and impact and influential process variables (n = 23)
Impact Variables
DASS Anxiety
DASS Depression
IES-R Total Score
PTV
AAS Positive Dimensions
AAS Negative Dimensions
PTGI Total Score
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
r
r2
-0.038
-0.015
-0.019
-0.014
-0.075
0.248
0.346
0.001
0.000
0.000
0.000
0.006
0.062
0.120
0.332
0.102
-0.012
0.053
0.110
0.010
0.000
0.003
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale.
The reverberation of this issue with children of survivors can be assessed by a
comparison of children of survivors who were born in Europe before their survivor parent/s
resettled and children of survivors who were born after their survivor parent/s emigration.
As Table 13.20 shows, there are no statistically significant differences between children of
survivors born in Europe before their survivor parents’ emigration and children of survivors
born in their parents’ post-war settlement location.
Table 13.20.
Children of survivors born before their survivor parents’ emigration from Europe versus those born after on impact and
influential process variables
Impacts
DASS Depression
DASS Anxiety
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
Born before emigration – in
Europe
(n = 27)
Born after emigration – not in
Europe
(n = 41)
Significance test results
η2
6.30 (6.60)
2.48 (3.10)
9.69 (4.82)
41.74 (8.80)
13.15 (4.24)
8.48 (10.44)
4.13 (4.38)
11.39 (4.63)
39.38 (10.19)
15.73 (7.15)
t (65.93) = 1.06, p = 0.30
t (65.73) = 1.83, p = 0.07
t (66) = 1.46, p = 0.15
t (66) = 0.99, p = 0.33
t (65.42) = 1.87, p = 0.07
0.014
0.042
0.031
0.015
0.041
39.35 (5.70)
97.89 (15.78)
33.81 (8.08)
33.65 (6.47)
41.29 (8.48)
93.30 (17.55)
34.75 (6.27)
32.49 (7.96)
t (66) = 1.05, p = 0.30
t (66) = 1.10, p = 0.28
t (64) = 0.53, p = 0.60
t (64) = 0.62, p = 0.54
0.016
0.018
0.004
0.006
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale.
13.2.6.2. – Influence on children of survivor’s perception of their parents/family
environment.
In this section, differences in the perceptions held by children of survivors born before and
after their parents’ emigration from Europe of family interaction patterns are considered.
As is the case for the impact and influential process variables, there are no statistically
© Janine Lurie-Beck 2007
263
significant differences in perceptions of family interaction that relate to whether a child of
survivors was born before, or after, parental emigration (refer to Table 13.21).
Table 13.21.
Perceptions of children of survivors born before and after their survivor parents’ emigration from Europe on family
interaction variables
Survivor Mother
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about the
Holocaust
HCQ – Indirect communication about the
Holocaust
HCQ – Guilt-inducing communication about
the Holocaust
HCQ – Frequent and willing communication
about the Holocaust
Survivor Father
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about the
Holocaust
HCQ – Indirect communication about the
Holocaust
HCQ – Guilt-inducing communication about
the Holocaust
HCQ – Frequent and willing communication
about the Holocaust
Non parent specific family interaction
variables
HCQ – Non-verbal presence of the Holocaust
FES – Cohesion
FES – Expressiveness
Born before emigration – in
Europe
(n = 27)
Born after emigration – not in
Europe
(n = 41)
Significance Testing
η2
9.38 (5.71)
2.73 (3.67)
4.42 (4.60)
46.50 (15.36)
2.48 (0.99)
8.00 (4.74)
2.97 (4.20)
5.72 (5.12)
41.31 (14.07)
2.71 (1.00)
t (64) = 0.49, p = 0.63
t (64) = 0.36, p = 0.72
t (64) = 0.53, p = 0.60
t (64) = 0.55, p = 0.58
t (46) = 0.80, p = 0.43
0.018
0.001
0.018
0.031
0.014
1.88 (1.18)
1.62 (0.82)
t (53) = 0.97, p = 0.34
0.018
3.12 (1.77)
3.52 (1.94)
t (53) = 0.80, p = 0.43
0.012
9.88 (2.73)
8.52 (2.98)
t (52) = 1.74, p = 0.09
0.055
8.29 (5.35)
4.46 (4.36)
6.17 (4.77)
45.50 (15.50)
2.43 (0.99)
8.68 (4.90)
3.22 (3.85)
4.86 (4.09)
45.17 (14.46)
2.46 (0.89)
t (63) = 0.17, p = 0.87
t (63) = 1.05, p = 0.30
t (63) = 0.96, p = 0.34
t (62) = 0.11, p = 0.91
t (54) = 0.11, p = 0.91
0.001
0.023
0.021
0.000
0.000
1.35 (0.71)
1.69 (0.92)
t (54.73) = 1.62, p = 0.11
0.040
2.88 (1.23)
3.44 (1.83)
t (57.99) = 1.44, p = 0.15
0.030
7.92 (3.05)
9.43 (3.19)
t (59) = 1.84, p = 0.07
0.054
9.87 (2.80)
43.88 (19.99)
37.44 (18.10)
10.38 (2.99)
41.50 (24.20)
35.84 (18.44)
t (64) = 0.69, p = 0.49
t (64) = 0.42, p = 0.68
t (65) = 0.35, p = 0.73
0.007
0.003
0.002
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire, FES = Family Environment Scale
13.2.7. – Post-war Settlement Location of Survivors
The final survivor demographic variable assessed is their post-war settlement location.
Specifically, the issue of interest here is whether a survivor remained in Europe (and the
site of their traumatisation) or immigrated to a distant continent such as Australia or
America or moved to Israel (a country specifically associated with the reason for the
persecution – i.e., Judaism).
13.2.7.1. – Influence on survivor and descendant psychological health.
The ANOVA analyses assessing post-war settlement location among survivors considered
three settlement regions. These regions are Europe, Australia/New Zealand and America.
Only one survivor in the sample currently lives in Israel and so it was not valid to include
Israel in the analysis. Because age is statistically significantly related to current region of
© Janine Lurie-Beck 2007
264
settlement (see Section 13.1) ANCOVAs were also conducted with age as a covariate
where age correlates statistically significantly with a variable that differs statistically
significantly by settlement location.
As can be seen in Table 13.22, survivors who settled in Australia or New Zealand
suffer more from PTSD symptoms (as measured by the IES-R) than those who went to
America or stayed in Europe. The Australian/New Zealand group also report significantly
more usage of maladaptive coping strategies than those who settled in America.
Table 13.22.
Holocaust Survivor post-war settlement group scores on impact and influential process variables
Europe a
(n = 3)
America c
(n = 8)
Significance Test Results
η2
Impact Variables
DASS Anxiety
DASS Depression
IES-R Total Score
Australia/New
Zealand b
(n = 12)
3.33 (2.31)
6.00 (7.81)
1.28 (1.47) b
5.55 (7.51)
6.36 (6.04)
4.81 (2.76) a c
4.63 (5.53)
6.13 (7.43)
1.58 (1.96) b
0.016
0.001
0.354
0.277
PTV
AAS Positive Dimensions
AAS Negative Dimensions
6.50 (2.18)
41.00 (5.29)
13.00 (6.56)
11.71 (4.92)
34.78 (7.31)
12.00 (2.69)
10.88 (3.27)
41.00 (9.77)
11.50 (4.44)
F (2,19) = 0.15, p = 0.86
F (2,19) = 0.01, p = 0.99
F (2,20) = 5.49, p < 0.05.
F (2,19) = 3.64, p < 0.05 ANCOVA with age
in 1945 as covariate
F (2,20) = 1.86, p = 0.18
F (2,17) = 1.42, p = 0.27
PTGI Total Score
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS – Benevolence
WAS – Meaningfulness
48.50 (47.38)
59.73 (30.18)
51.63 (25.36)
F (2,18) = 0.24, p = 0.79
0.025
40.67 (8.74)
100.00 (14.73)
34.67 (4.51)
39.33 (1.53)
46.55 (9.11) c
99.00 (17.45)
29.08 (6.97)
29.17 (11.46)
37.13 (5.84) b
109.50 (9.32)
30.75 (7.40)
33.13 (5.14)
F (2,19) = 3.27, p = 0.060
F (2,19) = 1.26, p = 0.31
F (2,20) = 0.80, p = 0.47
0.256
0.117
0.074
0.141
Kruskal Wallis
Kruskal Wallis
χ 2 (2) = 0.23, p = 0.89
χ 2 (2) = 4.44, p = 0.11
0.157
0.143
0.017
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale. Lettered superscripts indicate which groups differ statistically significantly according to Tukey post-hoc analyses.
Two post-war settlement location groups could be compared within the child of
survivor sample derived for this study: Australia/New Zealand and America. Only one
child of survivor participant had grown up in Israel and so an Israeli group could not be
included. The current country of residence of the grandchildren of survivors is a relatively
accurate indication of their survivor grandparent’s chosen post-war settlement location. It
was possible to compare grandchildren whose survivor grandparents had settled in
Australia or New Zealand to those who settled in America. There were no grandchildren of
survivors in the sample currently residing in Europe and only one residing in Israel and so
these regions/countries could not be included in the analyses.
The pattern of statistically significant results for descendants of survivors is
somewhat contradictory. Among the children of survivors, those who live in America
score statistically significantly lower on positive attachment dimensions than those in
© Janine Lurie-Beck 2007
265
Australia or New Zealand, but among grandchildren the pattern is reversed with
grandchildren in Australia or New Zealand scoring statistically significantly lower than
their American counterparts. The other statistically significant finding is that grandchildren
in Australia and New Zealand have a statistically significantly weaker belief in world
benevolence than American grandchildren. There have been no other studies to compare
survivors or descendants from these countries/regions to compare these results to.
Table 13.23.
Children and grandchildren of survivors’ post-war settlement group scores on impact and influential process variables
Impact Variables
DASS Anxiety
Australia/New
Zealand
(n = 46)
Children of survivors
America
Significance
(n = 21)
Test Results
3.32 (4.01)
3.76 (4.02)
DASS Depression
7.05 (8.05)
8.86 (11.41)
PTV
10.05 (4.90)
11.95 (4.22)
AAS Positive
Dimensions
AAS Negative
Dimensions
Influential
Psychological
Processes
COPE Maladaptive
42.01 (8.51)
36.19 (11.03)
14.43 (5.82)
15.00 (7.25)
40.07 (6.68)
41.71 (9.26)
COPE Adaptive
96.09 (16.66)
93.83 (17.71)
WAS –
Benevolence
WAS –
Meaningfulness
34.73 (6.36)
33.86 (8.40)
32.94 (8.03)
33.14 (6.13)
η
2
Australia/New
Zealand
(n = 21)
Grandchildren of survivors
America
Significance
(n = 6)
Test Results
t (65) = 0.42,
p = 0.67
t (65) = 0.74,
p = 0.46
t (65) = 1.53,
p = 0.13
t (65) = 2.36,
p < 0.05
t (65) = 0.34,
p = 0.73
0.003
5.10 (5.18)
1.67 (2.58)
0.008
5.43 (5.42)
3.00 (4.15)
0.035
10.45 (3.96)
8.42 (3.98)
0.079
38.43 (10.65)
47.50 (5.68)
0.002
15.50 (5.16)
19.50 (5.96)
t (65) = 0.82,
p = 0.41
t (65) = 0.51,
p = 0.62
t (63) = 0.46,
p = 0.64
t (65) = 0.10,
p = 0.92
0.010
39.12 (6.34)
44.17 (12.83)
0.004
88.71 (17.34)
97.00 (19.15)
0.003
34.12 (6.45)
39.67 (5.16)
0.000
36.83 (11.79)
41.33 (5.01)
η2
t (25) = 1.55,
p = 0.13
t (25) = 1.01,
p = 0.32
t (25) = 1.11,
p = 0.28
t (25) = 1.99,
p = 0.058
t (25) = 1.62,
p = 0.12
0.088
t (5.72) = 0.93,
p = 0.39
t (25) = 1.01,
p = 0.32
t (25) = 1.93,
p = 0.065
t (20.52) = 1.37,
p = 0.19
0.068
0.039
0.047
0.137
0.095
0.039
0.130
0.031
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
13.2.7.2. – Influence on survivor descendants’ perception of their parents/family
environment.
Table 13.24 presents descriptive and inferential statistics assessing differences in
perceptions of family environment held by children and grandchildren of survivors relating
to their survivor parent or grandparents’ post-war settlement location.
There are no
statistically significant differences relating to perceptions of survivor or child of survivor
mothers in relation to post-war settlement. However, children of survivor fathers who
settled in Australia or New Zealand rate their survivor father as statistically significantly
colder and also statistically significantly more likely to use guilt-inducing communication
about their Holocaust experiences than those in America. There are no previous studies
that have assessed this issue to make a comparison of this finding to.
© Janine Lurie-Beck 2007
266
Table 13.24
Children and grandchildren of survivors’ perceptions of family interaction stratified by survivor post-war settlement location
Children of Survivor Perceptions
Australia/ New
America
Zealand (n = 46)
(n = 21)
PCS – Warmth
8.66 (5.36)
8.31 (5.01)
PCS – Coldness
3.10 (3.78)
PCS – Ambivalence
5.24 (4.74)
Significance Test Results
Grandchildren of Survivor Perceptions
η2
Australia/New Zealand
America
(n = 21)
(n = 6)
Significance Test Results
η2
t (52) = 0.21, p = 0.84
0.001
11.55 (4.70)
14.00 (2.65)
t (20) = 0.87, p = 0.40
0.036
2.31 (4.52)
t (52) = 0.63, p = 0.53
0.007
5.08 (5.47)
t (52) = 0.11, p = 0.92
0.000
1.53 (3.20)
0.67 (1.15)
t (20) = 0.45, p = 0.66
0.010
4.00 (5.27)
3.00 (5.20)
t (20) = 0.31, p = 0.76
0.005
49.52 (14.80)
44.67 (14.22)
t (20) = 0.53, p = 0.60
0.014
Survivor/Child of Survivor Mother
PAQ – Fostering of Autonomy
44.56 (14.76)
40.46 (15.32)
t (52) = 0.87, p = 0.39
0.014
HCQ – Affective communication about the Holocaust
2.46 (0.79)
2.91 (1.45)
t (11.87) = 0.99, p = 0.34
0.039
Not applicable
HCQ – Indirect communication about the Holocaust
1.78 (1.01)
1.69 (1.03)
t (52) = 0.27, p = 0.79
0.001
Not applicable
HCQ – Guilt-inducing communication about the Holocaust
3.44 (1.95)
3.00 (1.63)
t (52) = 0.73, p = 0.47
0.010
Not applicable
HCQ – Frequent and willing communication about the
8.83 (3.01)
10.08 (2.57)
t (51) = 1.31, p = 0.20
0.032
Not applicable
PCS – Warmth
7.93 (5.15)
9.68 (4.83)
t (58) = 1.25, p = 0.22
0.026
11.50 (5.23)
8.75 (5.06)
t (12) = 0.90, p = 0.39
0.063
PCS – Coldness
4.41 (4.38)
2.37 (2.97)
t (49.86) = 2.12, p < 0.05
0.055
3.10 (3.57)
3.00 (6.00)
t (12) = 0.04, p = 0.97
0.000
PCS – Ambivalence
5.98 (4.16)
4.37 (4.69)
t (58) = 1.34, p = 0.19
0.030
5.00 (4.62)
6.75 (5.85)
t (12) = 0.60, p = 0.56
0.029
45.24 (14.99)
44.83 (12.33)
t (57) = 0.94, p = 0.35
0.015
51.05 (12.29)
50.25 (8.06)
t (12) = 0.12, p = 0.91
0.001
HCQ – Affective communication about the Holocaust
2.42 (0.97)
2.47 (0.85)
t (53) = 0.20, p = 0.84
0.001
Not applicable
HCQ – Indirect communication about the Holocaust
1.59 (0.88)
1.53 (0.84)
t (56) = 0.26, p = 0.80
0.001
Not applicable
HCQ – Guilt-inducing communication about the Holocaust
3.50 (1.68)
2.63 (1.42)
t (41.36) = 2.06, p < 0.05
0.062
Not applicable
HCQ – Frequent and willing communication about the
8.54 (3.26)
9.37 (3.11)
t (58) = 0.93, p = 0.36
0.015
Not applicable
Holocaust
Survivor/Child of Survivor Father
PAQ – Fostering of Autonomy
Holocaust
Non parent specific family interaction variables
HCQ – Non-verbal presence of the Holocaust
10.06 (3.18)
10.35 (2.30)
t (63) = 0.37, p = 0.71
0.002
FES – Cohesion
42.76 (22.08)
40.85 (24.18)
t (63) = 0.31, p = 0.76
0.002
46.90 (21.50)
51.83 (26.34)
t (25) = 0.47, p = 0.64
0.009
FES – Expressiveness
36.88 (18.54)
36.27 (17.94)
t (64) = 0.13, p = 0.90
0.000
49.75 (17.68)
50.50 (9.07)
t (17.06) = 0.14, p = 0.89
0.000
Not applicable
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust Communication Questionnaire, FES = Family Environment Scale
© Janine Lurie-Beck 2007
267
13.3. – Summary and Conclusions
This chapter has examined the interplay between measured variables in the model of the
differential impact of Holocaust trauma across three generations (the three categories of
psychological impact variables, influential psychological process variables and family
interaction/mode of trauma transmission variables) and survivor demographic and
situational variables. Consistent with one of the main arguments of the current thesis, it
was found that numerous demographic variables statistically significantly related to
survivor and descendant scores on psychological impact variables. This suggests that
survivors and descendants are not a homogenous group but are made up of many
distinguishable sub-groups with varying levels of psychological impairments.
It is of interest not only to consider each demographic variable separately, but also
to determine some kind of rank ordering of importance for the demographic variables. To
this end, Tables 13.25 and 13.26 summarise the strength of effects found for each of the
survivor demographic variables by providing the average as well as highest eta-squareds
( η 2 ) and co-efficients of determination (r
2
) which both provide a measure of the
proportion of variation in a dependent variable accounted for.
As was mentioned in
Chapter Eleven, Section 11.4, ranking by effect sizes rather than statistical significance
levels provides a ranking unfettered by power problems inherent with small sample sizes.
Figure 13.2 adds to the ranking of the influential psychological processes and
family interaction variables considered in Chapter Twelve by adding a ranking of the
survivor demographics considered in this chapter. As can be seen from the average and
highest proportion of variance accounted for, as well as the number of statistically
significant results associated with each demographic variable, the three most influential
demographics for both survivors and their children are the nature of the survivors’
experiences, the survivors’ country of origin and the loss of family experienced by the
survivor during the Holocaust. In other words, three aspects of the narrative of a survivor’s
Holocaust experience are the most important in determining survivor and descendant scores
on psychological impact variables.
These variables relate directly to specific details of their traumatic experience. The
nature of their experiences refers to whether they went through the camp system or endured
time in hiding. A survivor’s country of origin (as previously discussed in Chapter Four,
Section 4.4) relates closely to the duration of their persecution and the speed with which
© Janine Lurie-Beck 2007
268
they progressed through the persecution “system”. Finally, the third ranked variable of loss
of family obviously relates to their experiences of separation from loved ones such as
parents, spouses, extended family members, and the bereavement over their loss. It is
interesting, given the importance placed on post-trauma milieu in the trauma literature, that
details of survivors’ post-war resettlement, such as the speed of their resettlement, and the
country or region they chose to settle in, are ranked lower than the three key elements of the
traumatic experience itself. It is also interesting to note, given the theoretical conjecture as
to the importance of a survivors’ age or developmental stage during the Holocaust, that
survivor age during the Holocaust (or in 1945) is only ranked at number five. The pattern
of post-trauma environment factors being less important than elements of the Holocaust
experience itself is exactly mirrored for the children of survivors.
Chapter Fourteen continues the examination of the role of demographic variables by
considering descendant demographics. The ranking of the importance of demographic
variables is then revised to include both survivor and descendant demographic variables.
© Janine Lurie-Beck 2007
269
Table 13.25
Average and highest proportions of variance accounted for by survivor demographic variables among survivor and descendant scores on psychological impact and influential
psychological process variables.
Survivors
Children of Survivors
Grandchildren of Survivors
Average
Highest
Number of
Average
Highest
Number of
Average
Highest
Number of
2
2
2
2
2
2
2
2
2
Statistically
Statistically
Statistically
η or r
η or
η or r
η or
η or r
η or
significant
significant
significant
r2
r2
r2
results
results
results
Age during the Holocaust
0.05
0.205
1
0.019 father
0.070
1 for father
Unable to test due to insufficient data
0.012 mother
0 for mother
Gender
0.025
0.153
1
0.09
0.166
1
Unable to test due to insufficient data
Nature of Holocaust experiences
0.189
0.383
6
0.054 for father
0.121
2 for father
Unable to test due to insufficient data
0.017 for mother
0 for mother
Loss of family during the Holocaust
Sole survivor 0.04
0.117
0
0.017 for father
0.181
0 for father
Unable to test due to insufficient data
Ever alone 0.06
0.320
1
0.042 for mother
2 for mother
Country of origin
0.216
0.609
3
0.075 for mother 0.249
2 for father
Unable to test due to insufficient data
0.080 for father
2 for mother
Post-war settlement location
0.111
0.277
2
0.016
0.079
1
0.075
0.137
2
Length of time before resettlement
0.028
0.120
0
0.020
0.042
0
Unable to test due to insufficient data
Table 13.26
Average and highest proportions of variance accounted for by survivor demographic variables among descendant of survivors’ perceptions on family interaction variables
Children of Survivors Perceptions of Family Environment
Grandchildren of Survivors Perceptions of Family Environment
Holocaust Survivor Parent or
Average
Highest
Number of Statistically
Average
Highest
Number of Statistically
Grandparent Variables
significant results
significant results
η 2 or r 2
η 2 or r 2
η 2 or r 2
η 2 or r 2
Age during the Holocaust
Gender
Type/nature of H experiences
Loss of family
Country of origin
Post-war settlement location
Length of time before
resettlement
© Janine Lurie-Beck 2007
0.002 for father
0.006 for mother
0.021
0.040 for father
0.039 for mother
0.038 for father
0.030 for mother
0.072 for father
0.058 for mother
0.016
0.018
0.063 for father
0.021 for mother
0.057
0.096 for father
0.144 for mother
0.114 for father
0.070 for mother
0.246 for father
0.241 for mother
0.062
0.055
0 for father
0 for mother
0
1 for father
1 for mother
4 for father
2 for mother
3 for father
1 for mother
2 for father
0
Unable to test due to insufficient data
Unable to test due to insufficient data
Unable to test due to insufficient data
Unable to test due to insufficient data
Unable to test due to insufficient data
0.017
0.063
Unable to test due to insufficient data
0
270
1st Generation
(Survivors)
2nd Generation
(Children of Survivors)
3rd Generation
(Grandchildren of
Survivors)
Psychological Impact
Variables
Ranking of Influential
Psychological Processes
Depression
Anxiety
Paranoia
PTSD Symptoms
Romantic Attachment
Dimensions
Post-traumatic growth
1.
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions

2.
3.
4.
1.

2.
3.
4.
Depression
Anxiety
Paranoia
Romantic Attachment
Dimensions
1.

2.
3.
4.
Maladaptive coping
strategies
Assumption of World
Benevolence
Assumption of World
Meaningfulness
Adaptive coping
Strategies
Maladaptive coping
strategies
Assumption of World
Benevolence
Assumption of World
Meaningfulness
Adaptive coping
Strategies
Maladaptive coping
strategies
Assumption of World
Benevolence
Assumption of World
Meaningfulness
Adaptive coping
strategies
Ranking of Modes of Intergenerational
Transmission of Trauma
Ranking of Holocaust Survivor Demographic Moderators
1.
2.
3.
4.
5.
6.
7.
1. Parent-Child Attachment (especially
maternal)
2. Family Cohesion
3. Communication about Holocaust
experiences (specifically via affective or
non-verbal modes)
4. Encouragement of Independence
(maternal)
5. General Family Communication
1. Parent-Child Attachment (especially
maternal)
2. Family Cohesion
3. Encouragement of Independence
(maternal)
4. General Family Communication
Nature of experiences
Country of origin
Loss of family
Post-war settlement location
Age during the Holocaust (in 1945)
Gender
Length of time before resettlement
1. Survivor parent country of origin
2. Survivor parent experiences during the Holocaust
3. Survivor parent loss of family
4. Survivor parent gender
5. Survivor parent post-war settlement location
6. Survivor Parent age during the Holocaust (in 1945)
7. Length of time before survivor parents’ resettlement after
the war
Figure 13.2. Ranking (from most important to least important) of Influential Psychological Processes, Family Interaction Variables/Proposed Modes of Trauma Transmission and Survivor
Demographic Moderators in terms of their relative importance in predicting scores on Psychological Impact Variables
© Janine Lurie-Beck 2007
271
Chapter Fourteen – Empirical Assessment of the Moderating Role of Descendant
Demographic Variables
This chapter continues the process begun in Chapter Thirteen: the testing of the role
demographic variables play in the model of the differential impact of Holocaust trauma
across three generations in terms of the three classes of psychological variables; namely
psychological impact, influential psychological process and family interaction/trauma
transmission mode variables. While Chapter Thirteen considered demographic variables
intrinsic to the survivor generation, the current chapter examines demographic variables
that are intrinsic to the descendant generations.
The background literature to these
demographics was outlined in Chapter Five.
The specific demographic variables to be tested have been added to the
representation of the model overleaf and have been bolded for ease of reference.
Grandchildren of survivors are analysed both in terms of the demographics relevant to their
own generation and those relevant to their parents’ generation.
As was the case for the
survivor demographics discussed in the previous chapter, the influence of a number of
ancestral demographic variables cannot be tested for all generations because of a lack of
data or prohibitive sample sizes. Such variables are noted at the relevant juncture in the
chapter.
© Janine Lurie-Beck 2007
272
Survivors
Psychological
Impacts of the
Holocaust
Depression
Anxiety
Paranoia
PTSD symptoms
Romantic
Attachment
Dimensions
• Post-traumatic
Growth
•
•
•
•
•
Ranking of Influential Psychological
Processes
1.
2.
3.
4.
Children of Survivors
1.
2.
•
•
•
•
Depression
Anxiety
Paranoia
Romantic
Attachment
Dimensions
3.
4.
Ranking of Modes of Intergenerational
Transmission
of Trauma
Maladaptive coping strategies
Assumption of World
Benevolence
Assumption of World
Meaningfulness
Adaptive coping Strategies
Maladaptive coping strategies
Assumption of World
Benevolence
Assumption of World
Meaningfulness
Adaptive coping Strategies
1.
2.
3.
Grand-children of Survivors
4.
•
•
•
•
Depression
Anxiety
Paranoia
Romantic
Attachment
Dimensions
1.
2.
3.
4.
Maladaptive coping strategies
Assumption of World
Benevolence
Assumption of World
Meaningfulness
Adaptive coping Strategies
5.
1.
2.
3.
4.
Parent-Child Attachment (especially
maternal)
Family Cohesion
Communication about Holocaust
experiences (specifically via affective or
non-verbal modes)
Encouragement of Independence
(maternal)
General Family Communication
Parent-Child Attachment
(especially maternal)
Family Cohesion
Encouragement of Independence
(maternal)
General Family Communication
Demographic Moderators
Ranking of Holocaust
Survivor Generation Demographics
1.
2.
3.
4.
5.
6.
7.
Nature of experiences
Country of origin
Loss of family
Post-war settlement location
Age during the Holocaust (in 1945)
Gender
Length of time before resettlement
1.
2.
Survivor parent country of origin
Survivor parent experiences during
the Holocaust
Survivor parent loss of family
Survivor parent gender
Survivor parent post-war settlement
location
Survivor Parent age during the
Holocaust (in 1945)
Length of time before survivor
parents’ resettlement after the war
3.
4.
5.
6.
7.
Children of Survivor
Generation
Grandchildren of
Survivor
Generation
• Number of
survivor parents
• Delay between
the end of the
war and their
birth
• Birth before or
after survivor
parent/s
emigration
• Birth order
• Gender
• Number of
survivor parents
• Delay between
the end of the
war and their
birth
• Birth before or
after survivor
parent/s
emigration
• Birth order
• Gender
• Number of
child of
survivor
parents
• Birth order
• Gender
Figure 14.1. Addition of Holocaust Survivor Descendant Demographic Moderators to the Test Model of the Differential Impact of Holocaust Trauma across Three Generations
© Janine Lurie-Beck 2007
273
14.1. – Demographic Variable Inter-relationships
As was undertaken for the survivor demographics, inter-relationships between descendant
demographic variables have been examined so that potential confounds could be identified.
Where statistically significant relationships between demographic variables exist, they
serve to inform the use of controlled statistical analysis such as ANCOVA or partial
correlations when considering relationships between demographic variables and model
variables.
14.1.1. – Child of Survivor Demographic Variable Inter-relationships
Time lapse between the end of the war and the birth of children of survivors is statistically
significantly related, and therefore confounded, with a number of variables.
These
statistically significant relationships are outlined in this section. All analyses with variables
statistically significantly related to time lapse are conducted as partial correlation analyses
(with time lapse partialled out) or ANCOVAs, with time lapse entered as a covariate to
control for this confound.
There is a statistically significant difference in time lapse in birth (F (2, 65) = 21.19,
p < 0.001) between children of survivors with one versus two survivor parents and also
within the children of one survivor group, depending on whether the survivor parent is their
mother or father. Specifically, the average time lapse for children with two survivor
parents is 5.67 years (SD = 4.57, n = 51), followed by children with a survivor mother with
10.40 years (SD = 9.34, n = 5), with children with a survivor father born after the longest
delay of 18.42 years (SD = 9.85, n = 12). The average age of survivor parents in 1945 is
also correlated statistically significantly to the time lapse in birth of children of survivors.
Specifically, age of survivor mother in 1945 is correlated with time lapse at r = - 0.65 (p <
0.001, n = 55) and age of survivor father in 1945 is correlated to time lapse at r = - 0.60 (p
< 0.001, n = 62).
The size of the time lapse or delay between the end of the war and birth of children
of survivors also differs statistically significantly, depending on whether the child was born
before or after their survivor parents emigrated from Europe (t (66) = 2.78, p < 0.01).
Children of survivors born before emigration were born an average of 5.19 years after 1945
(SD = 6.64, n = 27), while children born after emigration were born an average of 10.29
years after 1945 (SD = 7.89, n = 41).
© Janine Lurie-Beck 2007
The birth order position of children of survivors is
274
also unsurprisingly statistically significantly correlated to time lapse between the war and
their birth (r = 0.43, p < 0.001).
14.1.2. – Grandchild of Survivor Demographic Variable Inter-relationships
The problems of confounding demographic variables are less of an issue for the analyses of
the grandchildren of survivor sample. This is mainly because many of the demographic
variables, that are related strongly within the survivor and children of survivor samples,
cannot be assessed in relation to the grandchildren sample due to lack of data or small subgroup sample sizes.
The one obvious relationship between two demographic variables assessed in this
section that presents a potential confound is the relationship between the number of child of
survivor parents and the number of survivor grandparents. A participant with only one
child of survivor parent can have either one or two survivor grandparents, while a
participant whose parents are both children of survivors can have between two and four
survivor grandparents. This relationship is taken into account and noted in the assessment
of both these variables in relation to grandchildren’s scores.
14.2. – Moderating Influence of Child of Survivor Demographics
In this section, the demographic variables that relate to the children of survivor generation
are examined in terms of their influence on the measured variables in the model.
14.2.1. – Number of Holocaust Survivor Parents
14.2.1.1. – Influence on descendant psychological health.
Differences between children of survivors, related to the number of survivor parents they
have, were analysed and the results are presented in this section.
There is a strong
relationship between the number of survivor parents a child has and the time lapse between
the end of the war and their birth. All statistically significant t-test results were re-run as
ANCOVAs (with time lapse as a covariate) where time lapse between the end of the war
and children of survivor’s birth was also statistically significantly related to the variable in
question.
The results of this set of analyses are presented in Table 14.1. As can be seen,
there is only one statistically significant difference related to number of survivor parents.
The children with one survivor parent report statistically significantly higher vulnerability
than children with two survivor parents. While this is incongruous with the hypothesis that
children with two survivors will not fare as well as children with one survivor parent
(DH7), it is consistent with the meta-analytic finding that children with one survivor parent
© Janine Lurie-Beck 2007
275
scored higher on paranoia measures than children with two survivor parents (see Chapter
Nine, Section 9.2.1).
Table 14.1.
Children of one versus two Holocaust survivor parents’ scores on impact and influential process variables
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
One survivor parent
(n =17)
Two survivor parents
(n =51)
Significance Test Results
η2
4.47 (4.85)
6.79 (8.26)
12.88 (4.39)
3.15 (3.62)
7.88 (9.45)
9.99 (4.68)
0.021
0.003
0.071
0.023
39.56 (10.61)
16.76 (7.79)
40.57 (9.43)
14.02 (5.58)
t (66) = 1.20, p = 0.24
t (66) = 0.42, p = 0.67
t (66) = 2.24, p < 0.05,
ANCOVA with time lapse in birth since 1945 as
covariate F (1, 65) = 1.53, p = 0.221
t (66) = 0.37, p = 0.71
t (21.73) = 1.34, p = 0.19
43.35 (10.70)
94.94 (18.05)
35.06 (4.80)
35.29 (6.23)
39.58 (5.95)
95.18 (16.69)
34.14 (7.64)
32.13 (7.63)
t (19.40) = 1.39, p = 0.18
t (66) = 0.05, p = 0.96
t (64) = 0.46, p = 0.65
t (64) = 1.54, p = 0.13
0.048
0.000
0.003
0.036
0.002
0.037
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
14.2.1.2. – Influence on survivor descendants’ perception of their parents/family
environment.
The results in Table 14.2 suggest that negative perceptions of survivor parents are
heightened when both parents are survivors, as opposed to when just one parent, is a
survivor. Specifically, if we turn to the statistically significant results in Table 14.2, it
seems that survivor mothers and fathers in a two survivor parent families are perceived to
make more frequent use of guilt-inducing communication about their Holocaust
experiences than when the other parent is not a survivor. In addition, children in two
survivor parent families report a statistically significantly higher non-verbal presence of the
Holocaust than one-survivor parent families.
These results also somewhat mirror the
thoughts of Hafner (1968) who suggested the presence of other survivors within a
household may serve to compound the effects of the Holocaust on an individual (with each
survivor parent further reinforcing the symptoms and affected parenting approaches of the
other).
© Janine Lurie-Beck 2007
276
Table 14.2.
Children of one versus two Holocaust survivor parents’ perceptions of family interactions
One survivor
parent
Survivor Mother Specific Variables
HCQ – Guilt-inducing communication about the
Holocaust
HCQ – Indirect communication about the
Holocaust
HCQ – Affective communication about the
Holocaust
HCQ – Frequent and willing communication
about the Holocaust
Survivor Father Specific Variables
HCQ – Guilt-inducing communication about the
Holocaust
HCQ – Indirect communication about the
Holocaust
HCQ – Affective communication about the
Holocaust
HCQ – Frequent and willing communication
about the Holocaust
Non parent specific family interaction variables
FES Cohesion
FES Expressiveness
HCQ – Non-verbal presence of the Holocaust
Two survivor
parents
(n =51)
Significance Test Results
η2
3.44 (1.91)
t (38.53) = 4.02, p < 0.001
0.046
1.33 (0.52)
1.80 (1.03)
t (54) = 1.09, p = 0.28
0.021
2.00 (0.00)
2.61 (1.00)
t (46) = 0.60, p = 0.55
0.008
7.50 (2.59)
9.29 (2.93)
t (53) = 1.42, p = 0.16
0.037
(n =12)
2.42 (0.67)
3.42 (1.74)
t (47.65) = 3.16, p < 0.01
0.061
1.50 (0.80)
1.57 (0.88)
t (57) = 0.27, p = 0.79
0.001
2.58 (0.67)
2.41 (0.98)
t (54) = 0.58, p = 0.57
0.006
10.17 (2.44)
8.51 (3.30)
t (59) = 1.63, p = 0.11
0.043
(n =17)
44.33 (20.61)
40.89 (19.23)
8.89 (2.63)
41.20 (23.39)
34.56 (17.64)
10.62 (2.86)
t (65) = 0.50, p = 0.62
t (66) = 1.28, p = 0.21
t (65) = 2.24, p < 0.05,
ANCOVA with time lapse in birth since 1945 as
covariate F (1,65) = 4.08, p < 0.05
0.004
0.024
0.072
0.061
(n = 6)
2.17 (0.41)
Notes. HCQ = Holocaust Communication Questionnaire, FES = Family Environment Scale. Differences in perceptions of
survivor parents in one and two survivor parent dyads on parent specific variables not related to Holocaust
communication are presented in Section 14.2.1.3 within the context of comparing perceptions of survivor to non-survivor
parents.
14.2.1.3. – Perceptions of survivor versus non-survivor parents.
Previous research has not looked at the perceptions of non-survivor spouses of survivor
parents, or the non-survivor parent in a one survivor parent family. Analyses reported in
this section represent the first assessment of this issue.
Research that has compared
perceptions of survivor and non-survivor parents have taken the perceptions of parents from
control or comparison groups where neither parent was a survivor and has left views of the
non-survivor parent in one survivor parent families unaddressed.
The same lack of
assessment also applies for the next generation, with no data existing on views of nondescendant of survivor parents in single child of survivor parent families.
© Janine Lurie-Beck 2007
277
Table 14.3.
Mean differences in ratings of survivor versus non-survivor parents on parent-child attachment dimensions and parental
facilitation of independence
Mother
PCS – Warmth
PCS – Coldness
PCS –
Ambivalence
PAQ – Fostering of
Autonomy
Father
PCS – Warmth
PCS – Coldness
PCS –
Ambivalence
PAQ – Fostering of
Autonomy
Perception of Survivor
Parent (total COHS
sample) a
n = 56
8.58 (5.17)
2.81 (3.83) d
5.19 (4.89)
Perception of Survivor Parent
(among children with one survivor
parent) b
n=6
11.25 (6.40)
1.00 (1.41)
4.25 (6.65)
Perception of Survivor Parent
(among children with two survivor
parents) c
n = 51
8.35 (5.07)
2.96 (3.94)
5.27 (4.80)
Perception of NonSurvivor Parent d
43.73 (14.63) d
46.25 (21.96)
43.52 (14.18)
52.91 (8.68) a
n = 12
9.73 (4.86)
2.45 (2.54)
4.36 (4.11)
n = 51
8.38 (5.04)
3.94 (4.27)
5.60 (4.53)
n=6
11.25 (2.99)
3.00 (3.56)
4.25 (4.35)
49.82 (14.27)
44.13 (14.92)
53.00 (5.89)
n = 61
8.63 (5.00)
3.66 (4.02)
5.37 (4.44)
45.19 (14.85)
n = 12
10.73 (4.29)
0.91 (1.04) a
3.27 (3.44)
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire. Lettered
superscripts indicate which groups differ statistically significantly according to Tukey post-hoc analyses.
An interesting pattern in relation to perceptions of survivor and non-survivor
parents emerges when perusing Table 14.3. In column A, the perceptions held by children
of survivors of their survivor parent (irrespective of whether that parent is their only
survivor parent or if they have two survivor parents) are definitely universally more
negative than the views held by children with one survivor parent of their non-survivor
parent (column D). Indeed, survivor mothers in this categorisation are seen as statistically
significantly colder (t (60.05) = 3.34, p < 0.01) and less encouraging of their children’s
independence (t (23.35) = 2.88, p < 0.01). However, when this issue was examined a little
more deeply, a pattern emerges that suggests that when there are two survivor parents, both
parents are viewed more negatively than the survivor parent in single survivor parent
families.
Could it be the case that survivor parents are generally less facilitating of
independence and less warm, more cold and more ambivalent, but the presence of a nonsurvivor parent lessens the effect or the presence of another survivor parent enhances the
effect? This pattern mirrors that found for the Holocaust communication variables assessed
in the previous section.
14.2.2. – Time Lapse between the Holocaust and the Birth of Children of Survivors
14.2.2.1. – Influence on descendant psychological health.
Based on the results of meta-regressions (presented in Chapter Nine, Section 9.2.7) it has
been hypothesised that the shorter the timeframe between the end of the war and the birth
of a child of survivors, the worse they will fare (DH9). In order to analyse this issue with a
raw data sample (which was not possible within the meta-analyses), correlations were
© Janine Lurie-Beck 2007
278
calculated between child of survivors’ scores and the delay between the end of the war and
their birth and are presented in Table 14.4.
Surprisingly, the correlations that are
statistically significant are in the opposite direction to that hypothesised (and found with the
meta-regression).
Specifically both anxiety and vulnerability scores for children of
survivors correlate positively with delay in birth rather than negatively as hypothesised.
Table 14.4.
Correlations between children and grandchildren of survivors’ scores on impact and influential process variables with the
time lapse between the Holocaust and the birth of children of survivors
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Children of Survivors
Time lapse in birth
(n = 68)
r
r2
Grandchildren of Survivors
Time lapse in Father’s birth
Time lapse in Mother’s birth
(n =14)
(n =19)
r
r2
r
r2
0.27 *
-0.02
0.25 *
- 0.08
0.09
0.073
0.000
0.063
0.007
0.009
-0.243
-0.207
-0.421
0.395
0.134
0.059
0.043
0.177
0.156
0.018
0.007
-0.007
0.090
-0.274
0.084
0.000
0.000
0.008
0.075
0.007
0.21
0.10
0.13
0.17
0.043
0.010
0.017
0.029
0.076
0.265
0.141
0.324
0.006
0.070
0.020
0.105
-0.055
-0.223
-0.450 #
-0.179
0.003
0.050
0.203
0.032
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale,
* p < 0.05, # p < 0.10 (denoted only with n < 30)
However, it is interesting to note that a number of correlations among the
grandchildren’s generation are in the hypothesised direction (albeit not statistically
significantly). Specifically of note is the negative correlation between delay in child of
survivor fathers’ birth and grandchildren’s vulnerability scores and the positive correlation
with positive attachment dimensions. However, the negative correlation between delay in
mother’s birth and world benevolence again suggest the counter-intuitive increase in
negative effects with increasing delays. Clearly the influence of the delay in birth of
children of survivors on their own and their children’s psychological well-being is still an
area that needs further clarification.
14.2.2.2. – Influence on descendants’ perceptions of their parents/family
environment.
Given that it is hypothesised that more positive outcomes are associated with a longer postwar delay in the birth of children of survivors, it follows that children of survivors’
perceptions of their family environment will be more positive with a longer delay. In
addition, the flow on hypothesis would hold that children of survivors born after a longer
delay would be viewed more positively by their own children (the grandchildren of
© Janine Lurie-Beck 2007
279
survivors) than children of survivors born after a shorter delay. Table 14.5 presents the
correlation analyses designed to test this issue.
There are no statistically significant
correlations with children of survivors’ perceptions of their family environment, but there
are a number of sizeable correlations for the next generation which are in the hypothesised
direction. Interestingly, the correlations are much stronger when paternal delay in birth is
considered, rather than maternal delay in birth. Analyses suggest that up to a third of the
variation in perceptions of child of survivor fathers are related to the post-war delay in the
child of survivor fathers’ birth. Longer delays are associated with perceptions of increased
warmth and fostering of autonomy/independence and decreased coldness and ambivalence,
as well as increased family cohesion. A decrease in family communication/expressiveness
with shorter delays in child of survivor mothers is also somewhat apparent. A number of
these moderately sized correlations are not statistically significant because of the small
sample sizes involved, but their magnitude means they are worthy of mention.
Table 14.5.
Correlations between children and grandchildren of survivors’ scores on family interaction variables with the time lapse
between the Holocaust and the birth of children of survivors
Children of Survivors
Time lapse in birth
r
Gender specific family interaction
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
Survivor Mother
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about the Holocaust
HCQ – Indirect communication about the Holocaust
HCQ – Guilt-inducing communication about the Holocaust
HCQ – Frequent and willing communication about the Holocaust
(n =55)
0.122
0.018
0.031
-0.069
0.093
-0.207
0.076
-0.066
Survivor Father
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about the Holocaust
HCQ – Indirect communication about the Holocaust
HCQ – Guilt-inducing communication about the Holocaust
HCQ – Frequent and willing communication about the Holocaust
Family Interaction Variables (non-parent specific)
HCQ Non-verbal presence of the Holocaust
FES Cohesion
FES Expressiveness
(n =61)
0.055
-0.078
-0.041
0.083
0.212
-0.048
-0.070
0.183
(n =66)
-0.12
-0.01
0.23
r2
Not applicable
Grandchildren of Survivors
Time lapse in Father’s birth Time lapse in Mother’s birth
(n = 13)
(n =19)
r
r2
r
r2
0.477
-0.571 *
-0.518 #
0.582 *
0.228
0.326
0.268
0.339
-0.129
0.021
0.006
0.009
0.017
0.000
0.000
0.000
0.015
0.000
0.001
0.005
0.009
0.043
0.006
Not applicable
0.004
0.003
0.006
0.002
0.007
0.045
0.002
0.005
0.033
0.013
0.000
0.052
Not applicable
0.599 *
- 0.046
Not applicable
0.359
0.046
0.002
- 0.415 #
0.002
0.172
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire, FES = Family Environment Scale.
* p < 0.05, # p < 0.10 (denoted only when n < 30)
© Janine Lurie-Beck 2007
280
14.2.3. – Child of Survivor Gender
14.2.3.1. – Influence on descendant psychological health.
The hypothesised moderating role of child of survivor gender on child of survivors’ scores,
as well as relationships between variables in the model of the differential impact of
Holocaust trauma, are examined in this section. If the hypotheses about gender are to be
supported a more negative picture for females and children through the female line would
be expected.
Table 14.6 presents the mean scores and significance testing of daughters versus
sons of survivors on psychological impact variables. While none of the differences are
significant, it is notable that, overall, it is the daughters of survivors who score less
favourably than the sons of survivors. Turning to the grandchildren of survivor generation,
Table 14.6 also presents data for members of this generation with a child of survivor
mother only versus those with a child of survivor father only. Following on with the flow
on hypothesis, it is predicted that children of child of survivor mothers would fare less well
than children of child of survivor fathers.
A cursory look at the psychopathological
measures shows that the children of child of survivor fathers have mostly scored worse than
the children of child of survivor mothers. While this is contradictory to the hypotheses
about gender, it is somewhat consistent with the findings within the current study relating
to the influence of survivor parent gender among children of survivors. However the one
mean difference that reached statistical significance among the grandchildren generation
has those with a child of survivor father reporting more frequent usage of adaptive coping
strategies than those with a child of survivor mother.
© Janine Lurie-Beck 2007
281
Table 14.6.
Female versus male children of survivors’ scores on impact and influential process variables
Females
(n = 44)
Impact
Variables
DASS Anxiety
Children of Survivors
Males
Significance
(n =24)
Test Results
4.03 (4.46)
2.46 (2.62)
8.86 (9.62)
5.31 (7.78)
11.49 (4.69)
9.29 (4.60)
AAS Positive
Dimensions
AAS Negative
Dimensions
Influential
Psychological
Processes
COPE
Maladaptive
COPE Adaptive
40.02 (10.18)
40.85 (8.83)
15.34 (6.42)
13.54 (5.89)
40.93 (6.81)
39.76 (8.77)
96.77 (16.03)
92.11 (18.36)
WAS Benevolence
WAS Meaningfulness
33.90 (7.55)
35.21 (5.96)
31.87 (7.15)
34.83 (7.56)
DASS
Depression
PTV
η
2
Child of
survivor
mother only
(n =12)
Grandchildren of Survivors
Child of
Significance
survivor father
Test Results
only (n = 6)
t (65.57) =
1.83, p = 0.07
t (66) = 0.61,
p = 0.13
t (66) = 1.09,
p = 0.28
t (66) = 0.34,
p = 0.74
t (66) = 1.14,
p = 0.26
0.037
5.50 (5.95)
7.50 (8.78)
0.035
6.58 (5.02)
9.33 (14.61)
0.050
11.29 (4.21)
10.33 (6.38)
0.002
40.08 (9.71)
47.33 (8.73)
0.019
16.25 (5.38)
19.33 (3.67)
t (66) = 0.61,
p = 0.54
t (66) = 1.09,
p = 0.28
t (64) = 0.73,
p = 0.47
t (64) = 1.59,
p = 0.12
0.006
42.25 (8.36)
43.50 (9.09)
0.018
81.00 (16.93)
99.33 (8.96)
0.008
35.33 (4.12)
32.50 (11.06)
0.038
37.50 (12.30)
39.67 (4.59)
η2
t (16) = 0.58, p
= 0.57
t (5.60) = 0.45,
p = 0.67
t (16) = 0.38, p
= 0.71
t (16) = 1.54, p
= 0.14
t (16) = 1.26, p
= 0.23
0.020
t (16) = 0.29, p
= 0.78
t (15.84) =
3.00, p < 0.01
t (16) = 0.80, p
= 0.43
t (16) = 0.41, p
= 0.69
0.005
0.022
0.009
0.129
0.090
0.274
0.039
0.011
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
14.2.3.2. – Influence on descendants’ perceptions of their parents/family
environment.
Table 14.7 presents the data as to how male and female children of survivors’ perceptions
differ regarding their survivor parents and their family environment. As can be seen, none
of the differences reach statistical significance; however there is an overall pattern of
daughters of survivors having more negative perceptions than sons of survivors.
© Janine Lurie-Beck 2007
282
Table 14.7.
Female versus male child of survivor perceptions of their survivor parent/s on family interaction variables
Survivor Mother Specific Variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about the Holocaust
HCQ – Indirect communication about the Holocaust
HCQ – Guilt-inducing communication about the Holocaust
HCQ – Frequent and willing communication about the Holocaust
Survivor Father Specific Variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about the Holocaust
HCQ – Indirect communication about the Holocaust
HCQ – Guilt-inducing communication about the Holocaust
HCQ – Frequent and willing communication about the Holocaust
Non Parent Specific Family Interaction Variables
FES Cohesion
FES Expressiveness
HCQ Non-verbal presence of the Holocaust
Females
Males
(n =37)
7.95 (5.38)
3.43 (3.97)
5.54 (5.06)
43.24 (14.50)
2.47 (0.98)
1.68 (0.97)
3.24 (1.67)
8.68 (2.94)
(n =40)
8.30 (5.38)
4.05 (3.93)
5.40 (4.53)
45.72 (13.60)
2.46 (0.88)
1.48 (0.82)
3.25 (1.74)
8.58 (3.32)
(n =43)
40.98 (24.55)
34.68 (18.17)
10.55 (2.76)
(n =19)
9.58 (5.10)
2.11 (4.03)
4.37 (4.41)
43.89 (15.83)
2.87 (0.99)
1.89 (1.05)
3.42 (2.19)
9.94 (2.78)
(n =21)
8.95 (4.42)
3.05 (4.35)
5.33 (4.18)
44.52 (17.02)
2.42 (1.02)
1.74 (0.93)
3.15 (1.42)
9.33 (2.97)
(n =24)
43.96 (18.82)
39.08 (18.14)
9.46 (3.04)
Significance Test Results
η2
t (54) = 1.09, p = 0.28
t (54) = 1.18, p = 0.24
t (54) = 0.86, p = 0.40
t (54) = 0.15, p = 0.88
t (46) = 1.29, p = 0.20
t (54) = 0.78, p = 0.44
t (54) = 0.34, p = 0.74
t (53) = 1.53, p = 0.13
0.022
0.025
0.013
0.000
0.035
0.011
0.002
0.042
t (59) = 0.48, p = 0.64
t (59) = 0.91, p = 0.37
t (59) = 0.06, p = 0.96
t (58) = 0.30, p = 0.77
t (54) = 0.15, p = 0.88
t (57) = 1.10, p = 0.28
t (58) = 0.22, p = 0.83
t (59) = 0.88, p = 0.38
0.004
0.014
0.000
0.002
0.000
0.021
0.001
0.013
t (58.53) = 0.56, p = 0.58
t (66) = 0.96, p = 0.34
t (65) = 1.49, p = 0.14
0.004
0.014
0.033
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire, FES = Family Environment Scale
Table 14.8 compares the perceptions held by the grandchildren of survivor
generation of their child of survivor parents of their mother versus their father. This has
been done within the grandchildren subgroup for whom both parents are the child of
survivors. There are no statistically significant differences.
Table 14.8.
Mean differences in ratings of child-of-survivor mothers versus child-of-survivor fathers among grandchildren with two
child-of-survivor parents on parent-child attachment dimensions and parental facilitation of autonomy
Perception of Mother (n =10)
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
10.55 (5.95)
2.70 (4.14)
6.00 (6.11)
46.00 (15.09)
Notes. PCS = Parental Care-giving Style Questionnaire
Perception of Father (n =10)
10.00 (5.21)
3.40 (3.41)
6.10 (4.61)
48.15 (12.19)
Significance Test Results
η2
t (9) = 0.33, p = 0.75
t (9) = 0.69, p = 0.51
t (9) = 0.11, p = 0.92
t (9) = 0.49, p = 0.64
0.012
0.051
0.001
0.026
14.2.3.3. – Interaction between parent gender and descendant gender.
Mixed design ANOVAS were conducted to test for interactions between child of survivor
and survivor gender on child of survivor perceptions of survivor parents (with child of
survivor gender as a between subjects IV and survivor gender as a within subjects IV).
None of these analyses reach statistical significance. It is likely that low power (resulting
from small sample sizes, in particular n =17 for male children of survivors and the analysis
of two rather than one independent variable) plays a large role in this, as the results of
© Janine Lurie-Beck 2007
283
numerous other analyses (where only one independent variable was assessed) suggest a
differential parental gender influence. Issues relating to power in the analyses of data from
the study were mooted in Chapter Eleven, but are particularly relevant here where a more
complicated analysis (requiring a larger sample for adequate power) has been used.
To further test the possible moderating role of child of survivor gender, separate
correlation matrices were calculated for male and female children of survivors for scores on
psychological impact variables and influential psychological process variables with their
scores on the family interaction/modes of transmission variables. The hypothesis being
tested by these gender stratified correlation matrices is that the relationships with family
interaction variables will be stronger for female than male children of survivors. The
method for calculating whether the difference between two correlation co-efficients (in this
case between correlation co-efficients for males and correlation co-efficients for females)
was obtained from Howell (1992, p. 251). In the majority of cases, there are stronger
relationships between family interaction variables and impact and process variables for
daughters of survivors than sons of survivors, with 49% of the 171 correlations being in the
same direction and stronger for daughters than sons, 33% of the correlations in the same
direction and stronger for sons than daughters and 18% of the correlations in opposing
directions for sons and daughters. Ten of the son and daughter correlations are statistically
significantly different from each other and nine of these come from the 18% of correlations
with different signs. These are presented in Table 14.9. The results of these analyses show
that there are certainly grounds to contend that there is an interaction between parent and
child gender on family environment and transmission of Holocaust trauma.
Table 14.9.
Statistically significantly different correlations between impact and influential process variables and family interaction
variables when stratified by child of survivor gender
Family Interaction Variables
Survivor Father
PCS – Father Cold
HCQ – Guilt-inducing communication about the Holocaust by Father
HCQ – Frequent and willing communication about the Holocaust by Father
HCQ – Frequent and willing communication about the Holocaust by Father
HCQ – Frequent and willing communication about the Holocaust by Father
HCQ – Indirect communication about the Holocaust by Father
Mother
HCQ – Affective communication about the Holocaust by Mother
HCQ – Affective communication about the Holocaust by Mother
HCQ – Frequent and willing communication about the Holocaust by Mother
HCQ – Indirect communication about the Holocaust by Mother
Impact and Influential Process Variables
Female
(n =32)
Male
(n =17)
PTV
WAS - Meaningfulness
AAS Negative Dimensions
AAS Positive Dimensions
WAS - Benevolence
AAS Positive Dimensions
0.30
0.18
-0.21
0.44
0.25
0.47
-0.41
-0.62
0.42
-0.29
-0.57
-0.28
DASS - Anxiety
AAS Positive Dimensions
DASS - Anxiety
AAS Positive Dimensions
0.05
-0.07
0.27
0.25
0.64
-0.80
-0.49
-0.46
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire, FES = Family Environment Scale
© Janine Lurie-Beck 2007
284
14.2.4. – Child of Survivor Birth Order
14.2.4.1. – Influence on children of survivor psychological health.
The effect of child of survivor birth order on their psychological health is assessed in two
ways: correlations with birth order position and comparisons between only children and
those with siblings.
It should be noted that while there is a statistically significant
relationship between birth order and time lapse in birth of children of survivors, it was not
necessary to conduct any controlled analyses (either partial correlations or ANCOVAs), as
none of the variables that correlate with time lapse in birth are statistically significantly
related to birth order as well. Based on the review of the literature on this issue (see
Chapter Nine, Section 9.2.10), it was hypothesised that middle order children would have
the highest scores on negative variables (DH8).
Table 14.10 presents the results of correlation and t-test analyses designed to test the
influence of birth order on children of survivors’ scores on impact and influential process
variables from the model. There are no statistically significant correlations with birth
order; however analyses comparing only children to those with siblings did yield one
statistically significant difference. Contrary to what has been hypothesised in the literature
(see Chapter Five, Section 5.2.3), but consistent with the results of studies that have
directly assessed birth order (see Chapter Nine, Section 9.2.10), it is the children of
survivors with siblings who rate themselves statistically significantly higher on depression
than only children.
Table 14.10.
Correlations with child of survivor birth order and differences between children of survivors who are only children and
children of survivors with siblings on impact and influential process variables
Correlations with birth
order (n =66)
r
r2
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
0.028
0.161
-0.022
0.005
0.001
0.026
0.000
-0.093
0.000
0.009
0.184
-0.084
-0.028
-0.192
0.034
0.007
0.001
0.037
Comparisons between only children and children with siblings
Only child
(n =14)
With siblings
(n =52)
Significance Test Results
η2
2.36 (2.56)
3.93 (3.41)
9.50 (4.70)
39.57 (5.43)
3.68 (4.28)
8.82 (9.99)
10.97 (4.83)
40.64 (10.43)
t (34.88) = 1.46, p = 0.15
t (60.37) = 2.95, p < 0.01
t (64) = 1.02, p = 0.31
t (41.28) = 0.52, p = 0.60
0.019
0.048
0.016
0.002
15.79 (6.66)
14.65 (6.18)
t (64) = 0.60, p = 0.55
0.006
40.07 (6.98)
92.79 (17.45)
34.50 (4.97)
34.43 (7.28)
40.72 (7.83)
95.41 (17.11)
34.00 (7.36)
32.29 (7.42)
t (64) = 0.28, p = 0.78
t (64) = 0.51, p = 0.61
t (62) = 0.24, p = 0.81
t (62) = 0.96, p = 0.34
0.001
0.004
0.001
0.015
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
© Janine Lurie-Beck 2007
285
14.2.4.2. – Influence on children of survivors’ perceptions of their parents/family
environment.
Table 14.11 reports the descriptive and inferential statistics examining the influence of
child of survivor birth order on their perceptions of their survivor parents and their family
of origin. There are only two statistically significant results, but both are consistent with
the findings of the previous section. Specifically, the statistically significant correlation
between birth order and perceptions of survivor fathers’ use of affective communication
about the Holocaust suggests that children of survivors, who are later in the birth order,
notice this type of paternal communication more frequently than those in higher birth order
positions. The other statistically significant result is also suggestive of more negative
perceptions among later born children, with family cohesion being rated statistically
significantly lower by children with siblings than only children.
Table 14.11.
Correlations with child of survivor birth order and differences between child of survivor only children and children of
survivors with siblings on family interaction variables
Correlations with Birth
Order
r
r2
Survivor Mother Specific Variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about
the Holocaust
HCQ – Indirect communication about the
Holocaust
HCQ – Guilt-inducing communication
about the Holocaust
HCQ – Frequent and willing
communication about the Holocaust
Survivor Father Specific Variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about
the Holocaust
HCQ – Indirect communication about the
Holocaust
HCQ – Guilt-inducing communication
about the Holocaust
HCQ – Frequent and willing
communication about the Holocaust
Family Interaction Variables (non-parent
specific)
HCQ Non-verbal presence of the
Holocaust
FES Cohesion
FES Expressiveness
(n =55)
-0.054
0.018
0.130
-0.120
0.137
-0.251
0.158
-0.176
(n =61)
-0.113
0.111
0.114
-0.141
0.368 **
-0.092
-0.003
0.076
(n =66)
0.003
0.000
0.017
0.014
0.019
0.063
0.025
0.031
0.013
0.012
0.013
0.020
0.135
0.009
0.000
0.006
Comparisons between only children and children with siblings
Only child
With siblings
Significance Test Results
η2
(n =12)
9.50 (4.93)
2.42 (4.52)
4.58 (3.70)
47.00 (15.60)
2.89 (1.05)
(n =43)
8.42 (5.33)
2.98 (3.79)
5.26 (5.19)
42.86 (14.61)
2.53 (0.98)
t (53) = 0.63, p = 0.53
t (53) = 0.43, p = 0.67
t (24.42) = 0.51, p = 0.62
t (53) = 0.86, p = 0.40
t (46) = 0.99, p = 0.33
0.007
0.004
0.003
0.014
0.021
2.17 (1.34)
1.63 (0.87)
t (53) = 1.67, p = 0.10
0.050
3.08 (2.11)
3.40 (1.80)
t (53) = 0.51, p = 0.61
0.005
9.50 (3.85)
9.05 (2.65)
t (52) = 0.47, p = 0.64
0.004
(n =11)
10.09 (4.46)
1.73 (2.76)
5.09 (2.59)
51.09 (13.66)
2.10 (0.99)
(n =48)
8.12 (5.18)
4.22 (4.20)
5.47 (4.74)
43.83 (14.92)
2.53 (0.90)
t (58) = 1.17, p = 0.25
t (58) = 1.88, p = 0.07
t (57) = 1.48, p = 0.15
t (27.53) = 0.37, p = 0.72
t (53) = 1.35, p = 0.18
0.023
0.057
0.001
0.037
0.033
1.64 (0.92)
1.51 (0.83)
t (56) = 0.44, p = 0.66
0.003
3.09 (1.45)
3.27 (1.69)
t (57) = 0.33, p = 0.75
0.002
8.55 (3.45)
8.86 (3.20)
t (58) = 0.29, p = 0.77
0.001
(n =14)
(n =52)
0.115
0.013
9.07 (3.05)
10.42 (2.83)
t (63) = 1.56, p = 0.13
0.037
-0.117
-0.083
0.014
0.007
51.43 (15.97)
44.21 (19.31)
39.63 (23.56)
34.42 (17.67)
t (30.36) = 2.19, p < 0.05
t (64) = 1.81, p = 0.08
0.047
0.048
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire, FES = Family Environment Scale.
** p < 0.01.
© Janine Lurie-Beck 2007
286
It was not possible to assess the impact of child of survivor birth order on grandchildren of
survivors due to lack of data.
14.3. – Moderating Influence of Grandchild of Survivor Demographics
In this section, the demographic variables from the model that directly pertain to the
grandchild of survivor generation are analysed in terms of their role in influencing scores
on impact, psychological process and family interaction variables.
14.3.1. – Number of Child of Survivor Parents/Survivor Grandparents
It was hypothesised that grandchildren of survivors’ scores on negative psychological
variables will increase with the number of ancestors affected by the Holocaust (DH7).
Grandchildren of Holocaust survivors have the possibility of having up to four ancestors
traumatised by the Holocaust. It was necessary to not only consider whether a grandchild
had one or two children of survivor parents but also the number of survivor grandparents
they had. It is possible that a grandchild of survivors with one child of survivor parent
could have one or two survivor grandparents, while grandchildren with two child of
survivor parents could have between two and four survivor grandparents. Given how
strongly related these two variables (number of survivor grandparents and number of child
of survivor parents) are, they are considered together in this section.
In both cases,
controlled analyses are conducted with ANCOVAs comparing grandchildren with one
versus two child of survivor parents (with number of survivor grandparents as a covariate)
and partial correlations with number of survivor grandparents (with number of child of
survivor parents partialled out).
14.3.1.1. – Influence on descendant psychological health.
Table 14.12 presents the results of ANCOVAs and partial correlations designed to partition
the unique contribution of number of child of survivor parents and number of survivor
grandparents on grandchildren of survivors.
As can be seen, there are numerous
statistically significant relationships with both variables. With respect to number of child of
survivor parents first, this variable statistically significantly relates to anxiety, depression,
vulnerability and belief in world benevolence, even when number of survivor grandparents
is partialled out. Counter-intuitively, it is the grandchildren with only one child of survivor
parent who are scoring statistically significantly less favourably here, than those whose
parents are both children of survivors. It is interesting to note, however, that statistically
© Janine Lurie-Beck 2007
287
significant relationships between number of child of survivor parents and coping styles are
not statistically significant when the number of survivor grandparents is partialled out.
Conversely, when attention is turned to the partial correlations examining the
unique relationships between number of survivor grandparents and grandchildren’s scores
(partialling out the number of child of survivor parents), there are a number of quite
sizeable relationships which are all in the hypothesised direction: namely that anxiety,
depression and vulnerability increase with increases in the number of survivor
grandparents, while belief in world benevolence weakens with an increase in the number of
survivor grandparents.
Table 14.12.
Children of one versus two child of Holocaust survivor parents’ scores on impact and influential process variables
Comparisons between children with one and two child of survivor parents
One child of
survivor parent
(n =18)
Significance Test Results
η2
Impact Variables
DASS Anxiety
Two child of
survivor
parents
(n =10)
6.17 (6.82)
2.90 (3.35)
DASS Depression
7.50 (9.00)
3.30 (4.03)
PTV
10.97 (4.86)
9.25 (3.12)
AAS Positive Dimensions
42.50 (9.79)
35.80 (10.44)
AAS Negative Dimensions
17.28 (4.99)
15.15 (6.19)
Influential Psychological
Processes
COPE Maladaptive
t (26) = 1.42, p = 0.17
F (1, 25) = 9.28, p < 0.01
t (26) = 1.39, p = 0.18
F (1, 25) = 11.64, p < 0.01
t (26) = 1.01, p = 0.32
F (1, 25) = 6.79, p < 0.05
t (26) = 1.70, p = 0.10
F (1, 25) = 0.00, p = 1.00
t (26) = 0.99, p = 0.33
F (1, 25) = 2.97, p = 0.10
0.071
0.271
0.069
0.318
0.037
0.214
0.100
0.000
0.036
0.106
42.67 (8.36)
37.05 (7.37)
COPE Adaptive
87.11 (16.98)
97.90 (17.14)
WAS - Benevolence
34.39 (6.99)
34.85 (9.18)
WAS - Meaningfulness
38.22 (10.25)
36.55 (11.69)
t (26) = 1.77, p = 0.09
F (1, 25) = 1.11, p = 0.30
t (26) = 1.61, p = 0.12
F (1, 25) = 0.01, p = 0.94
t (26) = 0.15, p = 0.88
F (1, 25) = 5.81, p < 0.05
t (26) = 0.39, p = 0.70
F (1, 25) = 0.11, p = 0.75
0.108
0.042
0.090
0.000
0.001
0.189
0.006
0.004
Correlations with number
of survivor grandparents
sr
sr2
0.463 *
0.214
0.517 **
0.267
0.430 *
0.185
- 0.196
0.038
0.270
0.073
0.040
0.002
0.166
0.028
- 0.475 *
0.226
0.031
0.001
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
* p < 0.05, ** p < 0.01.
14.3.1.2. – Influence on descendants’ perceptions of their family environment.t
Table 14.13 reports on how the number of child of survivor parents and survivor
grandparents impacts on grandchildren of survivors’ perception of their family
environment. There are no statistically significant results relating to family expressiveness,
but the results relating to family cohesion are somewhat confusing. They suggest that the
overlapping variance between number of child of survivor parents and number of survivor
© Janine Lurie-Beck 2007
288
grandparents is so large that it cannot be disentangled. What is clear, though, is that the
degree of survivor ancestry has a definite negative effect on the perception of family
cohesion among the grandchildren’s generation.
Table 14.13.
Children of one versus two Child of Holocaust survivor parents scores on impact and influential process variables
Comparisons between children with one and two child of survivor parents
FES Cohesion
FES
Expressiveness
One child of
survivor parent
(n =18)
55.44 (17.85)
Two child of
survivor parents
(n =10)
35.10 (23.04)
49.00 (16.25)
52.22 (15.08)
Significance Test Results
η2
t (26) = 2.61, p < 0.05
F (1,25) = 1.07, p = 0.31
t (25) = 0.50, p = 0.62
F (1,24) = 0.00, p = 0.97
0.207
0.041
0.010
0.000
Correlations with number of
survivor grandparents
sr
sr2
-0.051
0.003
0.071
0.005
Notes. FES = Family Environment Scale
Differences in perceptions of child-of-survivor parents in one and two child-of-survivor parent dyads on parent specific
variables are presented in Section 14.3.1.3 within the context of comparing perceptions of child-of-survivor to non-childof-survivor parents.
14.3.1.3. – Perceptions of child of survivor versus non-child of survivor parents.
Table 14.14 presents the mean rating of child of survivor and non-child of survivor parents
on parent-child attachment dimension as well as their perceived level of facilitation of
independence.
Table 14.14
Ratings of child of survivor versus non-child of survivor parents on parent-child attachment dimensions and parental
facilitation of independence
Perception of Child
Survivor Parent (total
grandchild sample) a
Mother
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of
Autonomy
Father
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of
Autonomy
Perception of Child of
Survivor Parent (among
those with two child of
survivor parents) c
n = 11
10.86 (5.74)
2.45 (4.01)
5.45 (6.07)
47.36 (15.01)
Perception of Non-Child
of Survivor Parent d
n = 22
11.89 (4.51)
1.41 (3.00)
3.86 (5.15) d
48.86 (14.89) d
Perception of Child
Survivor Parent (among
those with one child of
survivor parent) b
n = 12
13.00 (2.63)
0.33 (0.65)
2.08 (3.53)
51.25 (14.16)
n = 16
10.44 (5.02)
2.88 (3.93)
5.63 (4.80) d
50.41 (11.08)
n=6
11.17 (5.08)
2.00 (4.90)
4.83 (5.46)
54.17 (8.57)
n = 10
10.00 (5.21)
3.40 (3.41)
6.10 (4.61)
48.15 (12.19)
n = 10
11.90 (4.25)
1.10 (2.51)
2.20 (2.94) a
53.00 (11.84)
n=6
13.00 (2.97)
0.83 (1.60)
1.33 (2.07) a
58.33 (4.37) a
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire. Lettered
superscripts indicate which groups differ statistically significantly according to Tukey post-hoc analyses.
Almost the identical pattern to that seen when comparing perceptions of survivor
parents to non-survivor parents emerges when comparing perceptions of child of survivor
parents to non-child of survivor parents. Certainly the perceptions of child of survivor
parents of both genders (irrespective of whether the other parent is also a child of survivors
© Janine Lurie-Beck 2007
289
– reported in the second column/column “a”) are all more negative than the perceptions of
non-child of survivor parents. There are two statistically significant differences here on the
maternal side with child of survivor mothers seen as more ambivalent (t (21.56) = 1.83, p =
0.081) and less facilitating of independence (t (25.46) = 2.66, p < 0.05) than mothers who
are descended from survivors. In addition, child of survivor fathers (t (24.00) = 2.26, p <
0.05) are seen as more ambivalent than fathers with non-Holocaust traumatised parents.
The pattern of child of survivor parents being viewed more negatively when they
are a part of a child of survivor parent dyad, as opposed to part of a child of survivor and
non-child of survivor parent dyad, is practically identical to that seen for the preceding
generation. The question of whether the effects of having both parents with survivor
experience or ancestry appears to have more support for the affirmative.
14.3.2. – Grandchild of Survivor Gender
14.3.2.1. – Influence on descendant psychological health.
The hypothesis relating to gender is that female grandchildren of survivors will score
higher on negative variables and lower on positive variables than male grandchildren of
survivors. Table 14.15 displays the mean scores and significance tests examining gender
differences on impact and influential process variables among grandchildren of survivors.
As can be seen, females report statistically significantly higher depression, higher
attachment anxiety, higher usage of maladaptive coping strategies and a weaker belief in
world meaningfulness than male grandchildren. These results support the hypothesis and
are mostly consistent with meta-analytic results (see Chapter Nine, Section 9.2.9) as well as
gender differences in the general population (for example Oltmanns & Emery, 1995).
Table 14.15.
Female versus male grandchildren of survivors’ scores on impact and influential process variables
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Females
(n =20)
Males
(n =8)
Significance Test Results
η2
5.85 (6.03)
7.65 (8.60)
11.15 (4.54)
40.30 (10.23)
17.78 (5.44)
2.88 (5.59)
1.88 (2.36)
8.38 (3.20)
39.63 (11.39)
13.38 (4.21)
t (26) = 1.20, p = 0.24
t (26) = 1.85, p = 0.075
t (26) = 1.57, p = 0.13
t (26) = 1.85, p = 0.075
t (26) = 0.15, p = 0.88
0.053
0.117
0.086
0.001
0.139
43.93 (6.81)
89.60 (16.37)
34.48 (7.45)
35.23 (9.64)
32.50 (6.00)
94.38 (20.94)
34.75 (8.75)
43.63 (11.12)
t (26) = 4.14, p < 0.001
t (26) = 0.64, p = 0.53
t (26) = 0.08, p = 0.93
t (26) = 2.00, p = 0.056
0.397
0.016
0.000
0.133
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
© Janine Lurie-Beck 2007
290
14.3.2.2. – Influence on descendants’ perceptions of their parents/family
environment.
There are no statistically significant differences relating to grandchild of survivor gender in
terms of their perceptions of their child of survivor parents or family environment. Table
14.16 reports the descriptive statistics and the results of significance testing on this issue.
Table 14.16
Female versus male grandchildren of survivors’ perceptions of their child of survivor parents on family interaction
variables
Child of Survivor Mother Specific Variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
Child of Survivor Father Specific Variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
Non parent specific family interaction variables
FES Cohesion
FES Expressiveness
Females
Males
(n =15)
11.90 (4.47)
1.53 (3.09)
4.13 (5.36)
48.80 (13.11)
(n =11)
10.82 (4.77)
2.73 (3.93)
5.82 (5.17)
50.68 (8.60)
(n =20)
45.90 (23.33)
51.58 (14.60)
(n =7)
11.86 (4.95)
1.14 (3.02)
3.29 (5.02)
49.00 (18.26)
(n =5)
9.60 (6.02)
3.20 (4.38)
5.20 (4.38)
49.80 (16.57)
(n =8)
53.88 (17.53)
46.50 (18.47)
Significance Test Results
η2
t (20) = 0.02, p = 0.98
t (20) = 0.28, p = 0.78
t (20) = 0.35, p = 0.73
t (20) = 0.03, p = 0.98
0.000
0.004
0.006
0.000
t (14) = 0.44, p = 0.67
t (14) = 0.22, p = 0.83
t (14) = 0.23, p = 0.82
t (14) = 0.14, p = 0.89
0.013
0.003
0.004
0.001
t (26) = 0.87, p = 0.39
t (26) = 0.76, p = 0.45
0.028
0.023
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, FES = Family
Environment Scale
14.3.2.3. – Interaction between parent gender and descendant gender.
As was conducted for the children of survivors, separate correlation matrices were
calculated for male and female grandchildren of survivors for scores on psychological
impact variables and influential psychological process variables with their scores on the
family interaction/modes of transmission variables. The hypothesis being tested by these
gender-stratified correlation matrices is that the relationships with family interaction
variables will be stronger for female grandchildren than male grandchildren. The method
for calculating whether the difference between two correlation co-efficients (in this case
between correlation co-efficients for male grandchildren and correlation co-efficients for
female grandchildren) was obtained from Howell (1992, p. 251). In the majority of cases,
there are stronger relationships between family interaction variables and impact and process
variables for grandsons of survivors than granddaughters of survivors, with 52% of the 90
correlations being in the same direction and stronger for sons than daughters, 11% of the
correlations in the same direction and stronger for daughters than sons and 37% of the
correlations in opposing directions for sons and daughters. Three of the son and daughter
correlations are statistically significantly different from each other and two of these three
© Janine Lurie-Beck 2007
291
came from the 37% of correlations with different signs. These correlations are presented in
Table 14.17.
Table 14.17.
Statistically significantly different correlations between impact and influential process variables and family interaction
variables when stratified by grandchild of survivor gender
Family Interaction Variables
Impact and Influential Process Variables
Females
(n =20)
Males
(n =8)
PCS – Ambivalence – Paternal
PAQ – Fostering of Autonomy – Paternal
PAQ – Fostering of Autonomy – Paternal
AAS Negative Dimensions
AAS Negative Dimensions
AAS Positive Dimensions
0.08
0.17
-0.16
0.8
-0.69
0.83
Notes. AAS = Adult Attachment Scale, PCS = Parental Care-giving Style Questionnaire, PAQ = Parent Attachment
Questionnaire
14.3.3. – Grandchild of Survivor Birth Order
14.3.3.1. – Influence on descendant psychological health.
It was not possible to validly compare only child to non-only child grandchildren of
survivors because of insufficient samples (there is only one or two grandchildren who are
only children in each analysis); therefore only correlations with birth order are presented
here.
As can be seen the one statistically significant correlation with birth order among
grandchildren of survivors is the positive correlation with a belief in world meaningfulness.
This result suggests a strengthening of belief in world meaningfulness the younger or the
lower down the birth order the grandchild of survivors is. This result is consistent with the
hypothesis that higher birth order will be associated with more negative effects.
Table 14.18
Correlations with grandchild of survivor birth order and impact and influential process variables (n = 27)
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
r
r2
-0.087
-0.232
-0.203
-0.027
0.135
0.008
0.054
0.041
0.001
0.018
-0.046
-0.101
0.176
0.399 *
0.002
0.010
0.031
0.159
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
* p < 0.05.
© Janine Lurie-Beck 2007
292
14.3.3.2. – Influence on descendants’ perceptions of their parents/family
environment.
The way in which the birth order of grandchildren of survivors relates to their perceptions
of their parents (who are children of survivors) and family environment is considered in
Table 14.19. The statistically significant correlations all relate to perceptions of child of
survivor fathers. Specifically, later born grandchildren perceive their child of survivor
father as less warm, more ambivalent and less fostering of their autonomy/encouraging of
their independence than grandchildren born earlier in the birth order. Again, this result
points to more negative effects and perceptions being related to later, rather than earlier,
birth order positions and is consistent with the meta-analytic findings (Chapter Nine,
Section 9.2.10 in relation to children of survivors), but inconsistent with the anecdotal
literature (Chapter Five).
Table 14.19
Correlations with grandchild of survivor birth order and their perceptions of their child of survivor parents as rated on
family interaction variables
Child of Survivor Mother Specific Variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
Child of Survivor Father Specific Variables
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
Non parent specific family interaction variables
FES Cohesion
FES Expressiveness
r
(n =22)
-0.159
-0.054
0.094
0.018
(n =15)
-0.496 #
0.334
0.623 *
-0.422
(n =27)
0.051
-0.053
r2
0.026
0.003
0.009
0.000
0.246
0.112
0.388
0.178
0.003
0.003
Notes. PCS = Parental Care-giving Style Questionnaire, PAQ = Parental Attachment Questionnaire, FES = Family
Environment Scale
* p < 0.05, # p < 0.10 (denoted only when n < 30)
14.4. - Differences Related to Sample Characteristics
Differences among survivors and children and grandchildren of survivors related to their
membership of survivor or descendant organisations and their therapy history are assessed
in this subsection.
14.4.1. – Membership of Survivors Organisations/Support Groups
Following on from the results of the sub-set meta-analyses presented in Chapters Eight and
Nine that addressed this issue, it was hypothesised in Chapter Ten that survivors or
descendants who are members of a survivor or descendant organisation will display higher
© Janine Lurie-Beck 2007
293
levels of negative symptoms and lower levels of positive dimensions/variables than nonmembers (SH1). The results of analyses comparing members and non-members of survivor
organisations among the three generations in the study sample are presented in Table 14.20.
Among survivors, the one statistically significant difference finds a much higher
level of posttraumatic growth among members of survivor organisations than non-members
and also a lower usage of maladaptive coping strategies. Apart from anxiety, overall the
survivor organisation members score more positively than the non-members.
This is
inconsistent with both the hypothesis and the results of the meta-analyses on this issue (see
Chapter Eight, Section 8.9).
With respect to the children of survivors, those who are not members of a
descendant of survivor organisation report statistically significantly less usage of adaptive
coping strategies than children of survivors who have joined such organisations. There is
less of a discernible pattern overall for group membership among children of survivors.
For the grandchildren’s generation, the non-members score less favourably on a whole
range of variables from depression, anxiety and vulnerability to world assumption
variables, but the one statistically significant difference has non-members reporting
statistically significantly lower usage of maladaptive coping strategies than members. No
data exists prior to this study comparing grandchildren of survivors based on descendant
organisation membership status.
© Janine Lurie-Beck 2007
294
Table 14.20.
Comparison of survivor/descendant group member and non-members among survivors and descendants on impact and influential process variables
Survivors
Members
Non-Members
(n =9)
(n =13)
DASS Anxiety
6.22 (6.67)
4.15 (5.89)
DASS Depression
5.33 (5.79)
IES-R Total Score
2.64 (2.91)
PTV
AAS Positive Dimensions
Children of Survivors
Significance Test Results
Grandchildren of Survivors
Members
Non-Members
Significance Test
Members
Non-Members
(n =37)
(n =30)
Results
(n =3)
(n =23)
Significance Test Results
t (20) = 0.77, p = 0.45
3.47 (3.54)
3.28 (4.19)
t (65) = 0.19, p = 0.85
3.33 (2.89)
5.30 (6.44)
t (24) = 0.52, p = 0.61
6.54 (7.07)
t (20) = 0.42, p = 0.68
6.92 (7.29)
7.70 (10.15)
t (65) = 0.36, p = 0.72
4.00 (5.29)
6.17 (8.23)
t (24) = 0.44, p = 0.66
3.64 (2.83)
t (21) = 0.84, p = 0.41
9.36 (3.11)
11.83 (4.97)
t (18.66) = 1.44, p = 0.17
11.10 (4.88)
10.26 (4.65)
t (65) = 0.72, p = 0.47
6.67 (4.73)
10.87 (4.08)
t (24) = 1.65, p = 0.11
41.38 (9.30)
36.42 (7.49)
t (18) = 1.32, p = 0.20
40.78 (9.58)
39.81 (9.94)
t (65) = 0.41, p = 0.69
49.00 (6.08)
39.17 (10.77)
t (24) = 1.53, p = 0.14
15.50 (5.27)
14.14 (7.03)
t (65) = 0.88, p = 0.38
21.33 (4.51)
16.11 (5.39)
t (24) = 1.60, p = 0.12
Impact Variables
AAS Negative Dimensions
12.88 (5.06)
11.42 (3.00)
t (18) = 0.81, p = 0.43
PTGI Total Score
71.33 (12.41)
43.00 (30.76)
t (19) = 2.57, p < 0.05
Not applicable
Not applicable
Not applicable
Not applicable
Influential Psychological
Processes
COPE Maladaptive
38.33 (6.48)
45.23 (9.31)
t (20) = 1.92, p = 0.070
42.37 (6.90)
39.09 (7.85)
t (65) = 1.79, p = 0.08
52.67 (9.45)
38.80 (7.26)
t (24) = 3.02, p < 0.01
COPE Adaptive
99.94 (16.93)
102.27 (16.28)
t (20) = 1.92, p = 0.07
102.22 (16.36)
89.94 (15.30)
t (65) = 3.17, p < 0.01
104.67 (5.86)
89.57 (17.96)
t (24) = 1.42, p = 0.17
WAS - Benevolence
32.10 (6.54)
28.69 (7.32)
t (21) = 1.16, p = 0.26
33.93 (6.00)
34.84 (7.78)
t (63) = 0.51, p = 0.61
41.00 (5.57)
33.50 (7.86)
t (24) = 1.59, p = 0.13
WAS - Meaningfulness
33.70 (7.04)
30.92 (10.94)
t (21) = 0.70, p = 0.49
32.36 (7.03)
33.32 (7.79)
t (63) = 0.52, p = 0.61
44.67 (4.04)
36.41 (10.45)
t (24) = 1.33, p = 0.20
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events Scale – Revised, AAS = Adult Attachment Scale, PTGI =
Posttraumatic Growth Inventory, WAS = World Assumptions Scale
© Janine Lurie-Beck 2007
295
14.4.2. – Participation in Counselling/Therapy
A comparison is made in this section of study participants with, and without, a history of
some form or counselling or therapy. Among the Holocaust survivors, it seems that the
group of survivors who have never been in therapy score higher on all measures of negative
impacts (except for vulnerability and attachment anxiety). Survivors with no therapy
history report statistically significantly more PTSD symptoms (as measured by the IES-R).
but also report statistically significantly higher levels of posttraumatic growth and score
more favourably on both negative and positive attachment dimensions. Lower scores of
survivors who have been in therapy on negative impact variables may reflect improvements
in these symptoms post-therapy rather than lower levels existing prior to therapy.
Turning to the child of survivor generation, the opposite pattern to that noted for
survivors is apparent. Children of survivors who have been in some form of therapy score
statistically significantly higher on anxiety, depression and attachment anxiety and
statistically significantly lower on positive attachment dimensions.
In addition, more
frequent usage of maladaptive coping strategies is noted by those with therapy experience.
The pattern that emerges among the grandchildren of survivors is similar to that for
the survivors and opposite to that found for the children of survivors. Grandchildren who
have had some therapy score lower on negative symptom measures. In fact, they also have
much stronger beliefs in world meaningfulness and benevolence (statistically significantly
so for the latter).
However, the grandchildren with therapy history do score statistically
significantly higher on attachment anxiety than grandchildren with no therapy history.
The findings among survivors and grandchildren of survivors that those with
therapy experience score better than those with no therapy experience could be explained
by the benefits of the therapy itself.
Since no pre-therapy baseline measurement is
available, it is not possible to know whether these groups scored higher than the no-therapy
groups on negative symptom measures. The opposing results for the child of survivor
generation is more in keeping with what might be expected in a comparison of a clinical
and non-clinical sample. However, it also points to the potentially discomforting idea that
therapy has not been successful with this generation overall. Of course, it would be unwise
to make wide generalisations based on one study sample but these analyses do raise
important questions regarding gauging the effectiveness (via the collection of baseline data)
of therapy modes for survivors and descendants.
© Janine Lurie-Beck 2007
296
Table 14.21.
Comparison of survivor/descendant therapy history or no therapy history among Holocaust survivors and descendants on impact and influential process variables
Holocaust Survivors
Significance Test Results
Children of Survivors
Therapy
No Therapy (n =22)
Significance Test Results
Grandchildren of Survivors
Therapy
No Therapy
(n =7)
(n =16)
DASS Anxiety
3.57 (4.31)
5.67 (6.87)
t (20) = 0.74, p = 0.47
4.42 (4.32)
1.50 (2.04)
t (66.00) = 3.79, p < 0.001
3.53 (3.80)
7.00 (8.06)
t (13.27) = 1.32, p = 0.21
DASS Depression
5.00 (6.30)
7.07 (6.44)
t (20) = 0.71, p = 0.49
9.86 (9.95)
2.91 (4.35)
t (65.68) = 4.00, p < 0.001
4.80 (5.52)
8.27 (10.49)
t (24) = 1.10, p = 0.28
IES-R Total Score
1.42 (1.93)
4.04 (2.86)
t (21) = 2.20, p < 0.05
PTV
11.29 (4.18)
10.81 (4.52)
t (21) = 0.24, p = 0.82
11.48 (4.75)
9.11 (4.41)
t (66) = 1.96, p = 0.05
10.13 (4.18)
11.18 (4.77)
t (24) = 0.60, p = 0.56
AAS Positive Dimensions
33.00 (6.11)
40.38 (8.04)
t (18) = 2.11, p < 0.05
38.64 (10.03)
43.82 (7.96)
t (66) = 2.12, p < 0.05
40.67 (9.70)
41.64 (10.06)
t (24) = 0.25, p = 0.81
16.35 (6.51)
11.27 (3.97)
t (61.98) = 3.97, p < 0.001
18.43 (4.72)
12.64 (3.26)
t (24) = 3.50, p < 0.01
(n =46)
Therapy
No Therapy
(n =15)
(n =11)
Significance Test Results
Impact Variables
AAS Negative Dimensions
14.71 (4.31)
10.38 (2.69)
t (18) = 2.78, p < 0.05
PTGI Total Score
37.17 (31.81)
65.20 (23.00)
t (19) = 2.27, p < 0.05
Not applicable
Not applicable
Not applicable
Not applicable
Influential Psychological Processes
COPE Maladaptive
39.71 (6.07)
43.60 (9.80)
t (20) = 0.96, p = 0.35
42.72 (7.74)
35.92 (4.39)
t (63.93) = 4.61,p < 0.001
41.50 (8.95)
39.45 (8.26)
t (24) = 0.59, p = 0.56
COPE Adaptive
97.64 (14.46)
102.63 (16.92)
t (20) = 0.67, p = 0.51
96.01 (15.75)
93.26 (19.35)
t (66) = 0.63, p = 0.53
90.93 (16.81)
87.45 (18.14)
t (24) = 0.50, p = 0.62
WAS - Benevolence
29.86 (4.67)
29.81 (7.94)
t (21) = 0.01, p = 0.99
34.33 (6.55)
34.48 (8.05)
t (64) = 0.08, p = 0.94
37.83 (4.60)
30.82 (8.95)
t (24) = 2.61, p < 0.05
WAS - Meaningfulness
33.29 (6.34)
31.13 (10.37)
t (21) = 0.51, p = 0.62
33.08 (7.89)
32.67 (6.33)
t (64) = 0.21, p = 0.84
39.43 (7.71)
34.82 (12.57)
t (15.44) = 1.08, p = 0.30
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events Scale – Revised, AAS = Adult Attachment Scale, PTGI =
Posttraumatic Growth Inventory, WAS = World Assumptions Scale
© Janine Lurie-Beck 2007
297
14.5. – Summary and Conclusions
This chapter, in conjunction with Chapter Thirteen, has examined the interplay between
measured variables in the model of the differential impact of Holocaust trauma across three
generations (the three categories of psychological impact variables, influential
psychological process variables and family interaction/mode of trauma transmission
variables) and demographic and situational variables.
The demographic variables
considered were tracked not only for the generation to which they directly relate, but also
through the descendant subgroups they created. Consistent with one of the main arguments
of the current thesis, it was found that numerous demographic variables statistically
significantly related to survivor and descendant scores suggesting that survivors and
descendants are not a homogenous group, but are made up of many distinguishable subgroups with varying levels of psychological impairments and functioning.
Tables 14.22 and 14.23 act as a summary of the statistically significant effects of
descendant demographic variables on descendant scores on the model variables. As was
undertaken at the end of Chapter Thirteen, these tables report average, as well as highest
eta-squareds ( η 2 ) and co-efficients of determination (r2 ), which both provide a measure of
the proportion of variation in a dependent variable accounted for by demographic variables.
This process allows a rank ordering of the demographic variables in terms of their import to
the model.
In Figure 14.2, the ranking of both survivor demographics (considered in Chapter
Thirteen) and descendant demographics (discussed in this chapter) has been brought
together in order to provide overall rankings. What is startling to note is the continuing
importance of survivor parent demographic variables in determining the psychological
well-being of the children of survivor generation despite the addition of child of survivor
demographics. The three key Holocaust trauma elements of survivors’ country of origin,
the nature of their Holocaust experiences and the extent of familial losses which were the
most important predictors of survivor mental health, are reverberating through the
generations, and remain the top three demographic predictors among the children of
survivor generation, even with the addition of descendant demographics.
In fact,
demographics intrinsic to the child of survivor generation itself (such as gender, birth order
and post-war delay in birth) do not appear in the rankings until the very end of the ranked
list of demographic variables.
© Janine Lurie-Beck 2007
298
Indeed, ancestral demographic variables are also dominant for the grandchildren of
survivors with the number of child of survivor parents topping the ranking, and number of
survivor grandparents coming in in third position. Insufficient data was available to assess
the impact of many of the survivor generation demographics on the grandchild of survivor
generation. It would be interesting to see whether the three Holocaust trauma elements of
nature of experience, family loss and country of origin would continue to be as salient for
the grandchild of survivor generation as they are for both the survivor and child of survivor
generations. Despite this shortcoming, the fact that the extent of survivor ancestry a
grandchild of survivor has is the top ranking demographic variable in differentiating their
scores on psychological impact variables is suggestive of the continuing influence of the
Holocaust even two generations removed from direct suffering of the trauma itself.
It should be noted here that an attempt was made to conduct multiple regressions
with the demographic variables as predictors and psychological impact variables as
criterion/dependent variables in order to examine the unique and overlapping effects of the
demographic variables in a multivariate way with the children of survivors (for whom a
total sample size of 70 was deemed adequate for such an approach).
The approach
attempted was to dummy code each level of each categorical demographic variable (such as
nature of Holocaust experience). These analyses could not proceed due to a number of
reasons. The sample size was reduced substantially when ineligible cases/participants were
excluded (a reduction from the original 70 children of survivors to 32 children of survivors
when reduced to cases with no missing data for all the relevant demographic variables). A
number of the demographic variables could not be included because once the sample was
reduced to 32, there were insufficient or no participants in some levels/categories of the
dependent variable, rendering it meaningless to include in the analysis. This process
involved eliminating a number of demographic variables considered the most important in
the univariate approach (such as the nature of the survivor parents’ Holocaust experiences
and their country of origin). The small sample size also led to incredibly low power for the
analyses. In addition, further variables were eliminated because of collinearity problems.
These include predictably strong relationships between such as things as the birth country
of survivor mothers and survivor fathers. Because of these issues, a multivariate multiple
regression approach, while desirable and preferable to a univariate approach, was rendered
invalid. However, the reader should bear in mind that this approach was attempted.
© Janine Lurie-Beck 2007
299
Chapter Fifteen goes on to present a number of case studies from the survivor, child
of survivor and grandchild of survivor generations. The key elements of Holocaust trauma
(nature of experiences, family loss and country of origin) that have proved to be so pivotal
for the psychological health of survivors and their descendants are highlighted in the choice
of case studies. The case studies demonstrate how varied the experiences of Holocaust
survivors were during the war with a richness of detail that only case studies can provide.
© Janine Lurie-Beck 2007
300
Table 14.22
Average and highest proportions of variance accounted for by descendant demographic variables among descendant scores on psychological impact and influential psychological process
variables.
Children of survivors
Average
η
2
or r
Grandchildren of survivors
Highest
2
η
2
or r
2
Number of
Average
Statistically
η
significant results
2
or r
Highest
2
η
2
or r
Number of Statistically
2
significant results
Child of survivor
demographics
Number of survivor parents
0.019
0.048
1
0.115
0.267
Delay between the end of
0.028
0.073
2
0.073 for father
0.177 for father
0.042 for mother
0.203 for mother
the war and birth
Birth order
Gender
0.013 birth order
0.037 birth order
0
0.012 only child versus
0.048 only child versus with
0
with siblings
siblings
0.024
0.038
Unable to test
0.067
0.274
1
Not applicable
0.127
0.318
4
Birth order
Not applicable
0.036
0.159
1
Gender
Not applicable
0.104
0.397
4
Grandchild of survivor
0
4
Not applicable
demographics
Number of child of survivor
parents
© Janine Lurie-Beck 2007
301
Table 14.23
Average and highest proportions of variance accounted for by descendant demographic variables among descendant of survivors’ perceptions on family interaction variables
Children of survivors
Average
η2
or r
Grandchildren of survivors
Highest
2
η2
or r
2
Number of Statistically
Average
significant results
η2
or r
Highest
2
η2
or r
Number of Statistically
2
significant results
Child of survivor
demographics
Number of survivor parents
0.028
0.061
3
0.004
0.005
0
Delay between the end of
0.013
0.052
0
0.254 for father
0.359 for father
4 for father
0.032 for mother
0.172 for mother
0 for mother
the war and birth
Birth order
Gender
0.022 for birth order
0.063 for birth order
0 for birth order
0.021 for only child
0.057 for only child
1 for only child versus
versus with siblings
versus with siblings
with siblings
0.013
0.042
0
Unable to test
0.023
0.051
1
Not applicable
0.021
0.041
0
Birth order
Not applicable
0.097
0.388
2
Gender
Not applicable
0.011
0.031
0
Grandchild of survivor
demographics
Number of child of survivor
parents
© Janine Lurie-Beck 2007
302
Psychological
Impacts of the
Holocaust
Depression
Anxiety
Paranoia
PTSD symptoms
Romantic
Attachment
Dimensions
• Post-traumatic
Growth
3rd Generation
(Grand-children of Survivors)
2nd Generation
(Children of Survivors)
1st Generation
(Survivors)
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Depression
Anxiety
Paranoia
Romantic
Attachment
Dimensions
Depression
Anxiety
Paranoia
Romantic
Attachment
Dimensions
Ranking of Influential
Psychological Processes
Ranking of Modes of
Intergenerational Transmission of
Trauma
(1) Maladaptive coping
strategies
(2) Assumption of World
Benevolence
(3) Assumption of World
Meaningfulness
(4) Adaptive coping
Strategies
Holocaust
Survivor Generation
Children of Survivor
Generation
Grandchildren of Survivor
Generation
(1) Nature of experiences
(2) Country of origin
(3) Loss of family during the Holocaust
(4) Post-war settlement location
(5) Age during the Holocaust (in 1945)
(6) Gender
(7) Length of time before resettlement
(1) Maladaptive coping
strategies
(2) Assumption of World
Benevolence
(3) Assumption of World
Meaningfulness
(4) Adaptive coping
Strategies
(1) Maladaptive coping
strategies
(2) Assumption of World
Benevolence
(3) Assumption of World
Meaningfulness
(4) Adaptive coping
Strategies
Ranking of Demographic Moderators
(1) Parent-Child Attachment
(especially maternal)
(2) Family Cohesion
(3) Communication about Holocaust
experiences (specifically via
affective or non-verbal modes)
(4) Encouragement of
Independence (maternal)
(5) General Family Communication
(1) Parent country of origin
(2) Parent nature of Holocaust experiences
(3) Parent loss of family during the Holocaust
(4) Survivor Parent gender
(5) Parent post-war settlement location
(7) Parent age during the Holocaust (in 1945)
(8) Number of survivor parents
(6) Post-war delay in
birth
(9) Birth order
(10) Birth before or after
parent settlement
outside of Europe
(11) Gender
(3) Number of survivor grandparents
(4) Grandparent post-war settlement location
(1) Number of child of
survivor parents
(6) Delay between the end
of the war and the birth of
parents
(5) Parent gender
(7) Birth order
(2) Gender
(1) Parent-Child Attachment
(especially maternal)
(2) Family Cohesion
(3) Encouragement of
Independence (maternal)
(4) General Family Communication
Figure 14.2. Ranking (from most important to least important) of Influential Psychological Processes, Family Interaction Variables/Proposed Modes of Trauma Transmission and
Demographic Moderators (both survivor and descendant) in terms of their relative importance in predicting scores on Psychological Impact Variables
© Janine Lurie-Beck 2007
303
Chapter Fifteen – Holocaust Survivor and Descendant Case Studies
Chapters Thirteen and Fourteen (from empirical study data) as well as Chapters Seven,
Eight and Nine (based on meta-analyses) have demonstrated that to attempt to represent the
Holocaust survivor and descendant populations as homogeneous groups seriously
misrepresents the effects of the Holocaust. Not only are the range of Holocaust experiences
heterogeneous but so too are the post-Holocaust adjustment levels and trauma transmission
to subsequent generations.
The results of the empirical study have served to highlight a number of
demographic variables which appear to be of import in determining post-Holocaust mental
health outcomes. In this chapter, a number of case studies representing subgroups of these
key demographics will be described and discussed. They will highlight the wide range in
experience and psychological reactions to direct survival of the Holocaust or being
descended from a survivor. All participants’ names have been changed to protect their
identity.
A number of study participants have been selected from the sample to be
highlighted as case study examples.
They were chosen to reflect a diverse range of
Holocaust experiences and ancestral backgrounds and serve to emphasise the heterogeneity
of the influence of the Holocaust that has been argued throughout the current thesis. In
particular, demographic variables that are the most discriminatory in relation to
psychological impact variables are highlighted further by these case studies. Case studies
from each of the three generations are presented in turn, with comparisons made to relevant
sub-groups of the empirical study sample.
15.1. – Survivor Case Studies
Five Holocaust survivor case studies are presented in this section. The five cases were
chosen to highlight the heterogeneity of experiences a Holocaust survivor may have
endured during the Holocaust. They serve to further illuminate the influence of the key
demographic variables identified in Chapter Thirteen, namely nature of experiences, loss of
family and country of origin.
15.1.1. – “Zosia”- Polish Child Survivor who was in Hiding
Zosia is of Polish/Lithuanian descent. She was a young child during the Holocaust, turning
10 years old in 1945 at the end of the war. During the initial years of the war she remained
© Janine Lurie-Beck 2007
304
with her parents, as well as during their transfer to a Lithuanian ghetto. In 1941, Zosia was
smuggled out of the ghetto to live with another family under an assumed identity. She was
not reunited with her parents until after the war. Zosia immigrated to America in 1951 at
age 16, after spending time in both Austria and Italy in DP camps. In her current life, Zosia
is heavily involved in survivor and child survivor organisations, but reported she has never
attended any therapy. She married and had two sons and is now divorced. Zosia has
obtained tertiary qualifications.
Table 15.1 outlines Zosia’s scores on the psychological impact and influential
psychological process variables from the model of Holocaust trauma developed over the
course of the current thesis. As can be seen, Zosia’s scores indicate that she suffers from a
moderate level of anxiety and a mild level of depression. Her scores on these two measures
are much higher than the grand mean for the survivor sample, as well as sub-group means
for other survivors in hiding and other child survivors. Her scores for PTSD symptoms, as
measured by the Impact of Events Scale – Revised (IES-R) indicate that while her overall
score and scores on the intrusion and avoidance subscales are lower than the grand survivor
means and equivalent or lower than the relevant sub-group means, she has a higher score on
hyperarousal than both the grand and subgroup means. While a score of 1 for hyperarousal
on a scale from 0-4 hardly indicates severe pathology, it is noteworthy that she contrasts to
the rest of her cohort on this scale.
Zosia scores very high on post-traumatic growth, much higher than the grand mean
and both of the relevant sub-groups used for comparison. When her scores on the subscales of this measure are consulted it seems that she rates the identification of new
possibilities, personal strength and relating to others as being the most resultant of her
Holocaust experiences.
In terms of attachment, Zosia’s classification is on the cusp of the secure and
dismissing attachment types. This is because her rating of her comfort with being to close
to others is relatively high, but her rating of her comfort with depending on others is lower
than most others classified as secure.
© Janine Lurie-Beck 2007
305
Table 15.1.
Zosia’s scores compared to whole survivor sample and relevant survivor subgroups means on psychological impact and
influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
IES-R – Intrusion
IES-R – Avoidance
IES-R – Hyperarousal
IES-R – Total Score
PTV
AAS Positive
Dimensions
AAS Negative
Dimensions
AAS Type
PTGI Total Score
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Zosia
All
Survivors
Survivors in
Hiding
Children < 13
years in 1945
Minimum
Score
Maximum
Score
Highest Score in
Normal Range
12.00 (Moderate)
10.00 (Mild)
0.75
0.25
1.00
2.00
11.00
36.00
4.87
6.13
1.26
1.11
0.85
3.22
10.83
38.29
5.50
6.25
0.70
0.47
0.54
1.79
9.13
37.88
5.38
6.25
0.72
0.45
0.58
1.76
10.31
38.63
0.00
0.00
0.00
0.00
0.00
0.00
0.00
12.00
42.00
42.00
4.00
4.00
4.00
12.00
24.00
60.00
7.00
9.00
See note under
table.
16.00
12.00
14.63
13.25
6.00
30.00
Dismissing/Secure
71.00
56.09
49.00
55.00
0.00
105.00
43.00
118.00
32.00
37.00
42.22
101.61
30.04
32.08
41.63
102.50
33.75
38.25
40.13
103.31
33.25
38.63
24.00
36.00
8.00
12.00
80.00
144.00
48.00
72.00
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale. The scale authors suggest that scores on the IES-R be interpreted in terms of their position on the likert scale
used by participants. The scale points are denoted as follows in terms of the amount of distress caused in the past week
in relation to Holocaust experiences: 0 = not at all, 1 = a little bit, 2 = moderately, 3 = quite a bit, 4 = extremely.
15.1 2. – “Siegfried”- German Child Survivor who was in Hiding
Siegfried was also a child during the war turning 12 in 1945. Siegfried was born in the
Netherlands after his German parents fled Germany when Hitler came to power. He
remained with his parents, brother and some extended family until quite late in the war. In
1943, the family had a lucky escape when they were all arrested for deportation to
Auschwitz but were released after four or five days. However, in 1944 while most of his
family members were taken to concentration camps, Siegfried survived the remainder of
the war in hiding in the northern area of the Netherlands. While his parents, brother, an
uncle and his grandmother survived the war, an aunt and uncle and many family friends did
not survive the camps. Siegfried has remained in Europe since the end of the war,
remaining in the Netherlands until 1974 when he moved to Switzerland. In 2004, he
decided to move to Germany and while he lives very close to the Swiss border, he says that
his decision to move to Germany was a very hard and emotional one. Siegfried attained a
high school education. He married and had a son and a daughter but is now divorced. He
is not a member of a survivor organisation and has never been in any form of therapy.
Siegfried has written/published a book based on his Holocaust experiences.
© Janine Lurie-Beck 2007
306
While Siegfried’s scores for anxiety and depression are within the normal range and
his vulnerability score is also very low, his score for the PTSD symptom of intrusion is
quite high relative to the whole survivor sample and certainly to his relevant sample subgroups. His rating for intrusion suggests he suffers from this PTSD symptom to a moderate
extent. This finding suggests that Siegfried finds himself remembering or reliving his
Holocaust experiences a fair bit but his low scores on other pathology measures suggest
that he is not overly troubled by this.
Indeed, apart from the intrusion score, Siegfried scores remarkably well on all other
measures, certainly much better than the sample means would predict. For example, his
score for positive attachment dimensions is very high and for negative attachment
dimensions is very low. Siegfried’s coping strategies also appear to be predominantly
adaptive and healthy (low maladaptive coping score and high use of adaptive coping). His
scores for the two world assumptions variables are also suggestive of his robustness, with
his beliefs in world benevolence and meaningfulness quite strong in comparison to his
fellow survivors. In addition, Siegfried reports a very high post-traumatic growth score –
markedly higher than the grand and sub-group means quoted for comparison. The growth
factors he most attributes to his Holocaust experiences, in descending order, are personal
strength, relating to others and an appreciation of life.
Siegfried’s case study serves to highlight the fact that there are survivors who suffer
very few if any negative symptoms and have actually come out of their Holocaust
experiences perhaps psychologically stronger (refer to his scores in Table 15.2). It is
noteworthy that Siegfried rated the identification of his own personal strength as a result of
the Holocaust as the highest possible rating. It is quite possible that it was the fact that he
recognised in himself the strength to overcome and survive Nazi persecution that
emboldened him in his post-Holocaust years.
© Janine Lurie-Beck 2007
307
Table 15.2.
Siegfried’s scores compared to whole survivor sample and relevant survivor subgroups means on psychological impact
and influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
IES-R – Intrusion
IES-R – Avoidance
IES-R – Hyperarousal
IES-R – Total Score
PTV
AAS Positive Dimensions
AAS Negative
Dimensions
AAS Type
PTGI Total Score
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Siegfried
All
Survivors
Survivors in
Hiding
Children < 13 years
in 1945
Minimum
Score
Maximum
Score
Highest Score in
Normal Range
2.00
(normal)
2.00
(normal)
1.88
0.25
0.83
2.96
4.00
47.00
7.00
4.87
5.50
5.38
0.00
42.00
7.00
6.13
6.25
6.25
0.00
42.00
9.00
1.26
1.11
0.85
3.22
10.83
38.29
12.00
0.70
0.47
0.54
1.79
9.13
37.88
14.63
0.72
0.45
0.58
1.76
10.31
38.63
13.25
0.00
0.00
0.00
0.00
0.00
12.00
6.00
4.00
4.00
4.00
12.00
24.00
60.00
30.00
See note under table.
Secure
82.00
56.09
49.00
55.00
0.00
105.00
31.00
103.00
39.00
41.00
42.22
101.61
30.04
32.08
41.63
102.50
33.75
38.25
40.13
103.31
33.25
38.63
24.00
36.00
8.00
12.00
80.00
144.00
48.00
72.00
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale. The scale authors suggest that scores on the IES-R be interpreted in terms of their position on the likert scale
used by participants. The scale points are denoted as follows in terms of the amount of distress caused in the past week
in relation to Holocaust experiences: 0 = not at all, 1 = a little bit, 2 = moderately, 3 = quite a bit, 4 = extremely.
15.1.3. – “Greta”- Austrian Camp Survivor
According to the results of the empirical study conducted for the current thesis, Greta
suffered from the worst combination of experiences during the Holocaust. Born in Austria,
Greta was an adult during the Holocaust, turning 34 in 1945. In fact, she had married in
1936. Along with her husband and parents, Greta was forced to move into a ghetto in 1942
and was there for two years. In 1944, they were all sent to Auschwitz where her parents
and husband were killed in the gas chambers, making her the sole survivor of her family.
On her own from that point on, Greta moved through a number of labour and concentration
camps including Bergen-Belsen and Theriesenstadt. When liberated by the Russians in
1945 Greta was near death as a result of typhus.
After recuperating in Austria, she
immigrated to Australia in 1948. She remarried in 1950 and had a son before divorcing in
1973. In 1973, she married for a third time and was widowed in 2000. ‘Greta” is not a
member of a survivor organisation and has never had any form of therapy. She attained a
high school level education.
Not surprisingly, given her incredibly traumatic experiences during the Holocaust
(and also divorce and bereavement after the war), Greta suffers a fair amount from the
PTSD symptoms of intrusion and avoidance, and her scores are higher than the grand and
© Janine Lurie-Beck 2007
308
sub-group means quoted in Table 15.3. However, her hyperarousal score is very low,
especially in comparison to the camp survivor sub-group. Interestingly the mean for the
sole survivor group for hyperarousal is also quite low. Unfortunately there is some missing
data from Greta’s questionnaire booklet and so scores for vulnerability and posttraumatic
growth could not be obtained for her. On the page of the questionnaire booklet which
contained the PTGI, Greta noted that “it is very hard to imagine how I would have dealt
with life if I would not have been caught up in the Holocaust”.
However, ‘Greta” scores equivalently to her camp survivor peers on attachment
dimensions. She is classified as having a dismissing attachment type. Given the discourse
on this issue (see Chapter Two, Section 2.1 and Chapter Three, Section 3.1.1), it may be
ventured that a dismissing attachment type may have developed as a defence mechanism
following the overwhelming loss of family during the Holocaust.
Despite her elevated intrusion and avoidance scores, Greta also evidences some
resilience. Her scores on the world assumptions scales, while roughly equivalent to the
grand survivor mean, are much higher than her fellow camp survivors, suggesting that
despite her horrific experiences, her beliefs in world benevolence and meaningfulness were
not as shaken as they were for other survivors. She also scores quite highly on her use of
adaptive coping strategies.
Table 15.3.
Greta’s scores compared to whole survivor sample and relevant survivor subgroups means on psychological impact and
influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
IES-R – Intrusion
IES-R – Avoidance
IES-R – Hyperarousal
IES-R Total Score
PTV
AAS Positive Dimensions
AAS Negative Dimensions
AAS Type
PTGI Total Score
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Greta
All Survivors
Camp
Survivors
Sole
Survivors
Minimum
Score
Maximum
Score
Highest
Score in
Normal
Range
1 (normal)
1 (normal)
2.00
2.50
0.33
4.83
33.00
10.00
Dismissing
-
4.87
6.13
1.26
1.11
0.85
3.22
10.83
38.29
12.00
5.22
5.78
1.85
1.84
1.37
5.05
12.60
33.57
11.57
0.33
0.67
1.58
1.63
0.39
3.60
10.50
30.00
12.50
0.00
0.00
0.00
0.00
0.00
0.00
0.00
12.00
6.00
42.00
42.00
4.00
4.00
4.00
12.00
24.00
60.00
30.00
7.00
9.00
See note
under table.
56.09
68.11
56.00
0.00
105.00
43.00
112.00
32.00
30.00
42.22
101.61
30.04
32.08
46.78
96.89
26.40
27.30
40.33
96.67
29.33
25.00
24.00
36.00
8.00
12.00
80.00
144.00
48.00
72.00
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale. The scale authors suggest that scores on the IES-R be interpreted in terms of their position on the likert scale
used by participants. The scale points are denoted as follows in terms of the amount of distress caused in the past week
in relation to Holocaust experiences: 0 = not at all, 1 = a little bit, 2 = moderately, 3 = quite a bit, 4 = extremely.
© Janine Lurie-Beck 2007
309
15.1.4. – “Laszlo”- Hungarian Camp Survivor
Laszlo is a Hungarian born survivor. He was born in 1925, making him 20 years old at the
end of the war. While he notes that he endured numerous years of restrictions based on his
Judaism (and having to wear a yellow armband from 1940), he was not transferred to a
ghetto and it was not until 1944 that he was transported to a labour camp in Serbia. This
late transference to camps of Hungarian Jews corresponded with Hitler’s move to
exterminate Hungary’s Jews once the Hungarian government (once complicit) moved to
withdraw from its alliance with the Nazis.
Before this time, while the Hungarian
government had certainly enacted numerous laws which institutionalised anti-semitism, the
plan for mass extermination of Jews was not in force in Hungary. Indeed up until 1944,
Hungary was considered relatively safe for Jews and became a country to which to escape
(Edelheit & Edelheit, 1994).
Laszlo indicates that no family members died during the Holocaust, with both of his
parents and even his paternal grandmother managing to survive. After the war, Laszlo
spent a number of years in Belgium before immigrating to Australia in 1952 with his
parents. It is interesting to note that while he does not indicate whether his first wife was a
survivor or not, his second marriage in 1982 (after the death of his first wife in 1976 whom
he married in 1952) was to a fellow survivor from Hungary. Laszlo reports no membership
of any survivor organisation and also indicates no history of therapy.
Laszlo’s scores on impact and influential process variables in relation to the total
survivor sample and the camp survivor sub-sample are presented in Table 15.4. Laszlo
scores within the normal range for both anxiety and depression; in fact he indicates not
even the slightest experience of any symptoms associated with these two disorders. While
his cohort groups also score within the normal range, they do report much higher levels of
anxiety and depression symptoms. However, Laszlo reports quite frequent intrusion and
avoidance symptoms, certainly much higher than both the grand means for survivors and
the means for camp survivors. Interestingly, his hyperarousal score is relatively low being
equivalent to the survivor grand mean but notably less than the camp survivor mean.
Laszlo also scores relatively highly for posttraumatic vulnerability, indicating a notable
degree of fear about the safety of his loved ones and himself.
© Janine Lurie-Beck 2007
310
Table 15.4.
Laszlo’s scores compared to whole survivor sample and relevant survivor subgroups means on psychological impact and
influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
IES-R – Intrusion
IES-R – Avoidance
IES-R – Hyperarousal
IES-R Total Score
PTV
AAS Positive Dimensions
AAS Negative Dimensions
AAS Type
PTGI Total Score
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Laszlo
All Survivors
Camp
Survivors
Minimum
Score
Maximum
Score
Highest
Score in
Normal
Range
0.00 (normal)
0.00 (normal)
2.50
2.25
0.83
5.58
17.00
30.00
11.00
Dismissing
62.00
4.87
6.13
1.26
1.11
0.85
3.22
10.83
38.29
12.00
5.22
5.78
1.85
1.84
1.37
5.05
12.60
33.57
11.57
0.00
0.00
0.00
0.00
0.00
0.00
0.00
12.00
6.00
42.00
42.00
4.00
4.00
4.00
12.00
24.00
60.00
30.00
7.00
9.00
See note
under
table.
56.09
68.11
0.00
105.00
50.00
104.00
18.00
13.00
42.22
101.61
30.04
32.08
46.78
96.89
26.40
27.30
24.00
36.00
8.00
12.00
80.00
144.00
48.00
72.00
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale. The scale authors suggest that scores on the IES-R be interpreted in terms of their position on the likert scale
used by participants. The scale points are denoted as follows in terms of the amount of distress caused in the past week
in relation to Holocaust experiences: 0 = not at all, 1 = a little bit, 2 = moderately, 3 = quite a bit, 4 = extremely.
His scores are very low on the world assumptions scales, in particular world
meaningfulness. The lowness of these scores suggests he has little faith at all in the world
as a kind, caring place in which events are predictable and understandable. Perhaps given
these scores, it is not surprising that Laszlo is classified with a dismissing attachment style
– an understandable defence mechanism or survival adaptation.
Laszlo’s level of posttraumatic growth falls in the middle of the grand mean for
survivors and the mean for camp survivors. He indicates the growth factors most related to
his Holocaust experiences are an appreciation of life, personal strength and spiritual
change.
15.1.5. – “Hans”- German Survivor who Escaped in 1939
Hans was born in Germany in 1930 but was saved from the Holocaust via a kindertransport
to England in 1939. He indicates that that the only forms of persecution he directly
experienced prior to escaping from Germany was anti-semitism during a summer camp at
the age of six and being forced out of his school to go to a Jewish only school at age seven.
However, Hans’s father was taken to one of the first concentration camps in 1938 a year
before he escaped to England. During the war he stayed with an English family whom he
describes as “wonderful”. He was reunited with his parents in England in 1940 at which
point the family managed to immigrate to America. A brother also survived. While all
© Janine Lurie-Beck 2007
311
members of his family of origin not only survived the Holocaust but escaped from it, his
maternal grandparents were both killed during the war.
Hans describes his life as being “a succession of strokes of good luck”: getting a
spot on the kindertransport; escaping submarines on the trip across the Atlantic; and having
his parents also survive.
Because of this, Hans says he feels a special obligation to
tell/educate people about the Holocaust. To this end, Hans now gives talks to students from
primary school age to university.
He is also currently a member of two survivor
organisations, with one membership having lasted for the past 14 years. Hans did not
indicate any therapy history in his questionnaire.
Interestingly, while Hans obtains very low scores on the pathology measures for
PTSD symptoms and posttraumatic vulnerability compared to his cohorts, he scores much
higher on depression and anxiety (see Table 15.5). In fact, according to Lovibond and
Lovibond (1995), Hans can be described as suffering from mild levels of depressive
symptoms and moderate levels of anxiety symptoms.
Hans scores very favourably on the remainder of the impact and influential process
variables in the model of differential impact of Holocaust trauma.
Hans’s usage of
maladaptive coping strategies is much lower than the survivor mean and the escaped
survivor mean. His scores on the AAS attachment measure are almost identical to his
fellow escaped survivors’ means and he is classified with a secure attachment style. He
scores higher than both the grand mean for survivors and his fellow escaped survivors on
both of the world assumptions.
While his post-traumatic growth score is lower than the
grand mean for survivors, it is equivalent to that scored by other escaped survivors. He
rates his appreciation of life and his recognition of new possibilities as a result of the
Holocaust as being the most relevant growth factors for him.
© Janine Lurie-Beck 2007
312
Table 15.5.
“Hans’s” scores compared to whole survivor sample and relevant survivor subgroups means on psychological impact and
influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
IES-R – Intrusion
IES-R – Avoidance
IES-R – Hyperarousal
IES-R Total Score
PTV
AAS Positive Dimensions
AAS Negative Dimensions
AAS Type
PTGI Total Score
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Hans
All Survivors
Escaped
Minimum
Score
Maximum
Score
Highest
Score in
Normal
Range
14.00 (moderate)
13.00 (mild)
0.13
4.87
6.13
1.26
3.50
6.50
0.84
0.00
0.00
0.00
42.00
42.00
4.00
0.13
0.00
0.25
6.00
44.00
9.00
Secure
46.00
1.11
0.85
3.22
10.83
38.29
12.00
0.96
0.56
2.08
10.17
44.33
9.00
0.00
0.00
0.00
0.00
12.00
6.00
4.00
4.00
12.00
24.00
60.00
30.00
7.00
9.00
See note
under
table.
56.09
46.33
0.00
105.00
28.00
108.00
32.00
34.00
42.22
101.61
30.04
32.08
36.17
107.50
31.17
31.83
24.00
36.00
8.00
12.00
80.00
144.00
48.00
72.00
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale. The scale authors suggest that scores on the IES-R be interpreted in terms of their position on the likert scale
used by participants. The scale points are denoted as follows in terms of the amount of distress caused in the past week
in relation to Holocaust experiences: 0 = not at all, 1 = a little bit, 2 = moderately, 3 = quite a bit, 4 = extremely.
15.1.6. – Conclusions from Survivor Case Studies
While it cannot be denied that all five survivors, described in the case studies in the
previous sub-sections, suffered as a result of the Third Reich’s policy to persecute and
exterminate the Jewish population of Europe, it is clear that how an individual survivor
suffered as a result of this policy can differ widely. Table 15.6 provides a tabulated
comparison of the five case studies’ scores on the impact and influential process variables.
They are certainly heterogeneous. After reading the diverse narratives of the five survivors
and examining their scores, it is difficult to conceive why the dominant modus operandi of
survivor research has been to treat this population as if it was homogenous in its posttraumatic responses (see Chapter Two, Section 2.5.)
From a clinical perspective, it is interesting that the survivors who score the highest
on PTSD symptoms are the ones who have the lowest scores for anxiety and depression and
the survivors with the highest anxiety and depression scores evidence minimal PTSD
symptoms. Of course, this is just a sample of five survivors, but it is interesting to note the
apparent lack of co-morbidity of symptomatology.
© Janine Lurie-Beck 2007
313
In terms of country of origin, the fact that Hungarian survivors rated themselves the
least favourably in the empirical study conducted for the current thesis (see Chapter
Thirteen, Section 13.2.5.1) is reflected in the case of the Hungarian survivor Laszlo who
scores by far the lowest on the world assumptions scales and also the highest on the use of
maladaptive coping strategies. It is interesting that the survivor who may be labelled by
some as having suffered the most during the war, being a camp survivor and the sole
survivor of her family, Greta still has much stronger beliefs in world meaningfulness and
benevolence than Laszlo. Chapter Four, Section 4.4 explained the argument that survivors
who endured longer periods of persecution were potentially less affected than those with a
shorter period of persecution because of an acclimatisation process. Essentially, it was
argued that when survivors endured a number of years of restriction of liberties and then
ghettoisation before being transferred to camps, their experiences before they entered the
camps at least partially prepared them mentally for life in the camps. On the other hand,
survivors who were abruptly transported to camps with little if any acclimatisation stages,
were potentially more traumatised because the process for them was less gradual. This
argument is certainly supported when comparing Laszlo’s and Greta’s scores on the world
assumptions. Being Hungarian, Laszlo would have had a much more abrupt move towards
the camps than Greta who spent time in ghetto before her time in the camps.
Table 15.6.
Comparison of survivor case study scores on psychological impact and influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
IES-R – Intrusion
IES-R – Avoidance
IES-R – Hyperarousal
IES-R – Total Score
PTV
AAS Positive Dimensions
AAS Negative Dimensions
AAS Type
PTGI Total Score
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Zosia
Siegfried
Greta
Laszlo
Hans
12.00
(moderate)
10.00
(mild)
0.75
0.25
1.00
2.00
11.00
36.00
16.00
Dismissing/
Secure
71.00
2.00 (normal)
0.00 (normal)
14.00 (moderate)
0.00 (normal)
1.88
0.25
0.83
2.96
4.00
47.00
7.00
Secure
1.00
(normal)
1.00
(normal)
2.00
2.50
0.33
4.83
33.00
10.00
Dismissing
2.50
2.25
0.83
5.58
17.00
30.00
11.00
Dismissing
13.00
(mild)
0.13
0.13
0.00
0.25
6.00
44.00
9.00
Secure
82.00
-
62.00
46.00
43.00
118.00
32.00
37.00
31.00
103.00
39.00
41.00
43.00
112.00
32.00
30.00
50.00
104.00
18.00
13.00
28.00
108.00
32.00
34.00
2.00 (normal)
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, IES-R = Impact of Events
Scale – Revised, AAS = Adult Attachment Scale, PTGI = Posttraumatic Growth Inventory, WAS = World Assumptions
Scale
© Janine Lurie-Beck 2007
314
15.2. – Child of Survivor Case Studies
In this section, three child of survivor case studies are described. Each has a very different
Holocaust ancestry and the differences in impacts/transmission of Holocaust trauma can
clearly be seen. As was the case for the six survivor case studies presented, the child of
survivor cases chosen reflect the three most important demographic variables of survivor
parent’s type of Holocaust experiences, loss of family and country of origin. In addition,
differences related to perceived parental communication about the Holocaust, number of
survivor parents and post-war delay in birth or indirectly also highlighted.
15.2.1. – “Lena” – Daughter of two Polish Sole Survivors of the Camps
Both of Lena’s survivor parents were the sole surviving members of their families after the
Holocaust. Lena was born in 1956 in America, 11 years after the end of the war. Her
parents were both Polish Jews. Lena knows very few details of her father’s wartime
experiences, however she notes on her questionnaire that her mother was in hiding for over
a year before being caught trying to cross the Russian border. She then spent time in a
number of labour camps. She was transported (prior to Russia becoming one of the Allied
countries) deeper into Russia where she spent the rest of the war.
Lena has obtained tertiary qualifications.
She has never married and has no
children. Lena has participated in both individual and group therapy. She is a member of
several descendant organisations. On the page where participants were invited to write
down any thoughts, Lena provided a detailed outline of the efforts she has gone to to
document her parents’ Holocaust experiences, including piecing together some idea of her
father’s experiences which are largely unknown to her.
Table 15.7 reports Lena’s scores on the psychological impact and influential
psychological process variables from the model that are applicable to the child of survivor
generation. Lena’s scores are quite negative for the pathology measures. In fact, according
to the DASS cut-offs, Lena is classified as suffering from mild anxiety and extremely
severe depression. Interestingly though, other children of sole survivor mothers (whose
survivor fathers are not sole survivors) also score very highly on depression, although not
as highly as Lena has). However, given this, her vulnerability score is not as low as it
could be. Lena’s attachment scores are also far from positive and she is classified with a
fearful attachment style. With respect to coping strategies it appears that her usage of
adaptive coping strategies is lower than most of her cohort groups (except again for the
© Janine Lurie-Beck 2007
315
children with sole survivor mothers), and her usage of maladaptive coping strategies is
higher than the same comparison groups.
Given this largely negative picture it is
interesting to note that Lena’s scores for the world assumptions are comparatively high or
at least certainly not lower than the majority of the child of survivor means and sub-group
means presented in Table 15.7.
Table 15.7.
Lena’s scores compared to whole child of survivor sample and relevant child of survivor subgroups means on
psychological impact and influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive
Dimensions
AAS Negative
Dimensions
AAS Type
Influential
Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS Meaningfulness
Lena
All children
of survivors
All children of
two survivors
All children
of sole
survivor
father
All children
of
sole
survivor
mother
Minimum
Score
Maximum
Score
Highest
Score in
Normal
Range
9.00 (mild)
28.00
(extremely
severe)
8.00
25.00
7.50
3.43
3.15
7.88
4.67
11.08
4.00
23.00
0.00
0.00
42.00
42.00
7.00
9.00
10.66
40.32
9.99
40.57
13.08
36.83
6.33
30.33
0.00
12.00
24.00
60.00
27.00
14.71
14.02
18.25
16.33
6.00
30.00
40.52
95.12
34.38
33.08
39.58
95.18
34.14
32.13
42.00
97.25
31.83
33.33
42.67
80.33
36.00
29.33
24.00
36.00
8.00
12.00
80.00
144.00
48.00
72.00
Fearful
45.00
85.00
35.00
34.00
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
How Lena’s perceptions of her survivor mother and survivor father compare both to
each other and the total child of survivor sample is outlined in Table 15.8. Lena rates her
father as quite cold and her mother as very ambivalent in terms of their attachment
behaviour towards her. While she rates her father similar to the rest of the child of survivor
cohort in terms the degree to which he encouraged her autonomy, she rates her mother very
poorly in this regard.
With regard to her survivor parents’ communication about their Holocaust
experiences there is quite a stark difference. While she rates them similarly (and higher
than her cohort) on the level of guilt-inducing communication they used, she rates them as
polar opposites in terms of the frequency and willingness of their communication about
their experiences during the war. Lena notes that her mother spoke incessantly about her
experiences and was incredibly and perhaps over eager to do so. Her father, it seems, was
© Janine Lurie-Beck 2007
316
very reluctant to talk and she obtained very few details of his experiences. That a parent’s
silence with regard to their Holocaust experiences may lead to their children attempting to
imagine what happened (often worse than the reality) was alluded to in Chapter Three,
Section 3.4. Indeed, in the comments Lena made in her questionnaire booklet she hints that
this occurred for her. She imagined that his silence hinted that he had “experienced
something so mind blowing that it is literally unspeakable”. Lena theorises on the basis of
her research, with a few sparse details as her staring point, that he had a wife and family
and that he had quite possibly passed through a ghetto and then the Dachau concentration
camp where his family may have been killed. She believes that witnessing the death of
one’s wife and children would explain his strong reluctance to talk about his experiences.
Lena further commented on her attempts to get information out of her father: “Once I
pressured my mother to ask him about his past and he exploded in a rage. We got the
message and backed off.” These comments suggest that her father had not even shared
with her mother, his wife, what he had been through during the war.
Table 15.8.
Lena’s perceptions of her survivor father versus her perceptions of her survivor mother on gender specific family
interaction variables
Lena’s
perception of
her survivor
father
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about
the Holocaust
HCQ – Indirect communication about the
Holocaust
HCQ – Guilt-inducing communication
about the Holocaust
HCQ – Frequent and willing
communication about the Holocaust
Lena’s
perception of
her survivor
mother
2.00
7.00
4.00
47.00
-
All children of
survivors’
perceptions of
survivor
fathers
8.52
3.70
5.38
45.30
2.45
Minimum
Score
Maximum
Score
6.00
3.00
13.00
25.00
-
All children of
survivors’
perceptions of
survivor
mothers
8.50
2.98
5.14
43.46
2.59
0.00
0.00
0.00
14.00
1.00
16.00
16.00
16.00
70.00
5.00
1.00
1.56
3.00
1.75
1.00
5.00
6.00
3.22
6.00
3.30
2.00
10.00
3.00
8.84
15.00
9.09
3.00
15.00
Notes. PCS = Parental Caregiving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire
Given Lena’s rating of her parents on attachment and encouraging of autonomy
measures, it is perhaps unsurprising that she rates the level of cohesion in her family as
very low, much lower even than the children with a sole survivor mother which scores
much lower than the other cohort groups. The extent to which Lena felt a non-verbal
presence of the Holocaust when she was growing up is relatively comparable to other
children of survivors, particularly as it is elevated slightly among others who have at least
one sole survivor parent.
© Janine Lurie-Beck 2007
317
Table 15.9.
Lena’s perceptions of family interaction patterns compared to whole survivor sample and relevant survivor subgroups
means
FES – Cohesion
FES – Expressiveness
HCQ – Non-verbal presence
of the Holocaust
Lena
All children of
survivors
All children of two
survivors
9.00
34.00
12.00
42.04
36.23
10.16
41.20
34.56
10.62
All children of
sole survivor
father
44.33
40.58
11.91
All children of
sole survivor
mother
18.67
36.33
11.33
Minimum
Score
Maximum
Score
1.00
15.00
3.00
68.00
73.00
16.00
Notes. FES = Family Environment Scale, HCQ = Holocaust Communication Questionnaire
15.2.2. – “Otto” - Son of two Dutch Survivors who were in Hiding
Otto is the child of two Dutch survivors who as a married couple survived the war in
Holland under assumed non-Jewish identities. Otto was born at the very end of the war, a
matter of a few months before his parents were able to come out of hiding. He has a sister
who was born a number of years later in 1947. The family immigrated to New Zealand in
1950 before further immigrating to Australia in 1952.
Otto is divorced, but has a daughter and two sons. He has attained a tertiary
education. He also indicated that he had spent a year in Israel in 1971-1972. Otto has been
a member of a descendant organisation for over ten years and has had some individual
counselling. While he indicates that he has not had any group therapy, it is worthy of note
that the descendant organisation he is a member of meets regularly.
Otto has taken a strong interest in the Holocaust and in specifically recording his
family’s story. The family that helped his parents maintain their assumed identity during
the war also migrated to Australia and Otto has kept in touch with them and says he still
makes an annual visit to the surviving member of this couple. Perhaps because of the fact
that he was born while his parents were still in hiding, and his sister was born a number of
years later, Otto notes that his strong interest in the Holocaust, and the role it plays in
shaping his identity, is not shared by his sister at all. This point highlights the fact that
there can be differences even within survivor families, as to how the Holocaust experiences
of the parents can impact and influence their children.
In relation to his peers, and in general, Otto’s profile of scores is very positive (refer
to Table 15.10). He scores very low on anxiety and depression and vulnerability. His
attachment scores are suggestive of a very strongly secure attachment type. His usage of
adaptive coping strategies is much higher than all of the comparison data, although his
usage of maladaptive coping strategies (while equivalent to the whole generation sample) is
© Janine Lurie-Beck 2007
318
higher than children of survivor parents with like experiences. Otto’s belief in world
assumptions is also quite strong.
Table 15.10.
Otto’s scores compared to whole child of survivor sample and relevant child of survivor subgroups means on
psychological impact and influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive
Dimensions
AAS Negative
Dimensions
AAS Type
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Otto
All children of
survivors
All children of two
survivors
All children of
two hiding
parents
Minimum
Score
Maximum
Score
Highest Score in
Normal Range
1.00
(normal)
0.00
(normal)
5.00
57.00
7.50
3.15
2.00
0.00
42.00
7.00
3.43
7.88
3.40
0.00
42.00
9.00
10.66
40.32
9.99
40.57
6.50
48.60
0.00
12.00
24.00
60.00
14.00
14.71
14.02
12.20
6.00
30.00
40.52
95.12
34.38
33.08
39.58
95.18
34.14
32.13
35.87
105.40
38.20
34.80
24.00
36.00
8.00
12.00
80.00
144.00
48.00
72.00
Secure
40.00
110.00
39.00
38.00
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
Otto rates his mother very favourably (very warm and encouraging of his
developing autonomy) but his father not so (colder and ambivalent and discouraging of his
autonomy). In relation to their communication about the Holocaust with him, ‘Otto” rates
both his parents equivalently as relatively lower than his cohort on the negative modes of
communication and quite high on their willingness and frequency of discussions on the
subject (refer to Table 15.11).
Table 15.11.
Otto’s perceptions of his survivor father versus his perceptions of his survivor mother on gender specific family interaction
variables
Otto’s
perception of
his survivor
father
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about
the Holocaust
HCQ – Indirect communication about the
Holocaust
HCQ – Guilt-inducing communication
about the Holocaust
HCQ – Frequent and willing
communication about the Holocaust
Otto’s
perception of
his survivor
mother
5.00
8.00
11.00
25.00
2.00
All children of
survivors’
perceptions of
survivor
fathers
8.52
3.70
5.38
45.30
2.45
Minimum
Score
Maximum
Score
13.00
1.00
1.00
58.00
2.00
All children of
survivors’
perceptions of
survivor
mothers
8.50
2.98
5.14
43.46
2.59
0.00
0.00
0.00
14.00
1.00
16.00
16.00
16.00
70.00
5.00
1.00
1.56
1.00
1.75
1.00
5.00
3.00
3.22
3.00
3.30
2.00
10.00
11.00
8.84
10.00
9.09
3.00
15.00
Notes. PCS = Parental Caregiving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire
© Janine Lurie-Beck 2007
319
Otto rates the level of cohesion within his family as lower than other children of
survivors and much lower than other children of survivors with similar hiding experiences.
However, he does rate the level of expressiveness/general communicativeness within the
family as equivalent to the other cohorts and more positively rates the degree of non-verbal
presence of the Holocaust very low.
Table 15.12.
Otto’s perceptions of family interaction patterns compared to whole child of survivor sample and relevant child of survivor
subgroups means
Otto
FES – Cohesion
FES – Expressiveness
HCQ – Non-verbal presence of
the Holocaust
38.00
34.00
5.00
All children of
survivors
42.04
36.23
10.16
All children of two
survivors
41.20
34.56
10.62
All children of two
hiding parents
47.20
34.46
7.80
Minimum
Score
1.00
15.00
3.00
Maximum
Score
68.00
73.00
16.00
Notes. FES = Family Environment Scale, HCQ = Holocaust Communication Questionnaire
15.2.3. – “Mimi” - Daughter of a Female Belgian Child Survivor who was in Hiding
Mimi is a relatively young member of the child of survivor generation having been born in
1972 in England, (a post-war delay in birth of 27 years). However, this is understandable
when one discovers that her survivor mother was only five years old when the war ended in
1945. Mimi’s mother spent three years in hiding with an assumed identity in Belgium. Her
maternal grandparents were also in hiding close by and her mother was occasionally visited
by her own mother, in “health emergencies”. However, although she and her parents
survived, Mimi’s mother’s grandparents all died during the war. Her mother recalls that
she was very well treated by the family who looked after her during the war. After the war,
having been reunited with her parents, Mimi’s mother immigrated to Israel in 1950, but she
and her family returned to Belgium in 1952 due to financial difficulties. They attempted
emigration again in 1961 to England. This is where Mimi was born and where her mother
met her South African born father. Mimi’s family remained in England until 1994 when
they moved to Canada. In 2000, Mimi moved to New Zealand on her own. Mimi is the
last born child of three children. She has a tertiary education. She has never married.
Mimi indicates that she is not a member of a descendants’ organisation and she has never
had any therapy of any kind.
Table 15.13 compares Mimi’s scores on impact and influential process variables to
the whole children of survivor sample, as well as children of survivor mothers only. As can
be seen, Mimi scores very well overall, with normal and/or lower than her cohort’s scores
on depression, anxiety and vulnerability. She scores very favourably on the attachment
© Janine Lurie-Beck 2007
320
measure comfortably being assigned a secure attachment type. Her scores on coping
strategies are also comparatively favourable, with a very low maladaptive coping score and
very high adaptive coping score. Mimi’s belief in world benevolence is also quite strong
and much higher than the other relevant child of survivor groups; however her belief in
world meaningfulness is not as strong as her cohorts.
Table 15.13.
Mimi’s scores compared to whole child of survivor sample and relevant child of survivor subgroups means on
psychological impact and influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
AAS Type
Influential Psychological
Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Mimi
All children of
survivors
All children with a
survivor mother only
Minimum
Score
Maximum
Score
Highest Score in
Normal Range
1.00
(normal)
3.00
(normal)
8.00
51.00
6.00
Secure
7.50
1.20
0.00
42.00
7.00
3.43
4.70
0.00
42.00
9.00
10.66
40.32
14.71
11.00
42.90
12.00
0.00
12.00
6.00
24.00
60.00
30.00
40.52
95.12
34.38
33.08
38.20
87.80
32.60
32.60
24.00
36.00
8.00
12.00
80.00
144.00
48.00
72.00
32.00
111.00
38.00
26.00
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
Table 15.14 presents the comparison of Mimi’s perceptions of her parents to the
relevant whole sample perceptions. Mimi rates her survivor mother comparatively very
favourably: notably warmer and more encouraging of her autonomy. In terms of her
communication about her Holocaust experiences, Mimi’s perceptions match those of other
children of survivors.
Table 15.14.
Mimi’s perceptions of her survivor mother versus her non-survivor father r on gender specific family interaction variables
Mimi’s
perception of
her survivor
mother
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
HCQ – Affective communication about
the Holocaust
HCQ – Indirect communication about the
Holocaust
HCQ – Guilt-inducing communication
about the Holocaust
HCQ – Frequent and willing
communication about the Holocaust
16.00
1.00
0.00
65.00
2.00
All children of
survivors’
perceptions of
survivor
mothers
8.50
2.98
5.14
43.46
2.59
1.00
1.75
2.00
3.30
9.00
9.09
Mimi’s
perception of
her nonsurvivor father
15.00
1.00
1.00
59.00
All children of
survivors’
perceptions of
non-survivor
fathers
11.25
3.00
4.25
53.00
Not applicable
Minimum
Score
Maximum
Score
0.00
0.00
0.00
14.00
1.00
16.00
16.00
16.00
70.00
5.00
1.00
5.00
2.00
10.00
3.00
15.00
Notes. PCS = Parental Caregiving Style Questionnaire, PAQ = Parental Attachment Questionnaire, HCQ = Holocaust
Communication Questionnaire
© Janine Lurie-Beck 2007
321
Mimi’s overall perceptions of her family environment are considered in Table
15.15. While she notes a similar level of non-verbal presence of the Holocaust, Mimi rates
her family as notably more cohesive and expressive than other children of survivors do,
both overall and among children with a survivor mother only.
Table 15.15.
Mimi’s perceptions of family interaction patterns compared to whole child of survivor sample and relevant child of survivor
subgroups means
Mimi
FES – Cohesion
FES – Expressiveness
HCQ – Non-verbal presence of the Holocaust
60.00
60.00
9.00
All children of
survivors
42.04
36.23
10.16
All children with a survivor
mother only
49.17
38.33
8.50
Notes. FES = Family Environment Scale, HCQ = Holocaust Communication Questionnaire
Minimum
Score
1.00
15.00
3.00
Maximum
Score
68.00
73.00
16.00
15.2.4. – Summary and Conclusions from Child of Survivor Case Studies
The three child of survivor case studies have served to highlight a number of issues relevant
to the differential transmission of Holocaust trauma. Lena’s case emphasises the impact of
differing communication styles about Holocaust experiences. Her frustration with the lack
of information she obtained from her father led her to frequently try to imagine what he
went through and spend countless hours trying to establish the details of his experiences
through historical research.
In addition, further pointing to the heterogeneity of the survivor group, survivors
within a two survivor couple who experienced very similar/identical experiences during the
war can be perceived very differently by their children. Otto’s parents were in hiding
together during the war and yet Otto rates his father as cold and ambivalent and his mother
as warm. Of course, it cannot be determined if these differences would have been apparent
before the war/if the war had not occurred, but it is an interesting point to note.
Certainly there are differences between the three cases in terms of the level of
symptomatology they experience and the differences do reflect the results of the empirical
study. It is Lena, the child of two sole survivors of the Holocaust, who would be expected
to, and does, score the least favourably on the study measures.
15.3. – Grandchild of survivor case studies
Two grandchildren of survivor case studies are presented in this section. They were chosen
to represent the two extremes of degree of survivor ancestry with one having one survivor
grandparent and one having four survivor grandparents.
The number of survivor
grandparents and child of survivor parents were in the top ranked demographic variables in
© Janine Lurie-Beck 2007
322
influencing psychological impact variable scores among the grandchildren generation in the
empirical study. The overall theme that emerges from these two case studies is that no
matter what the extent of familial connection with the Holocaust is, it still leaves its mark in
some way.
15.3.1. – “Geena” - Grandchild with One Survivor Grandparent
Geena was born in 1984 and was aged 21 at the time of her participation in the current
study. She lives in America. Geena’s paternal grandmother is a Polish Holocaust survivor
who was aged 13 years at the end of the war. Her grandmother spent time in a ghetto but
escaped camp internment by living in hiding, initially with a false identity with another
family and then unaided. Geena’s father was born in the US in 1956 (a post-war delay in
birth of 11 years).
Her mother is also US born, but not descended from Holocaust
survivors. Geena reports that she first learned of her grandmother’s Holocaust experiences
at the age of five.
Geena has been a member of a descendant of survivor’s organisation for a number
of years. She also indicates that she has in the past had some individual counselling. Geena
has a tertiary qualification. She is single, but moved out of her family home at age 17.
Table 15.16 presents Geena’s scores on the impact and influential process variables
in relation to the whole grandchild of survivor sample and other grandchildren with one
survivor grandparent. Geena’s scores are very positive overall. Her scores on all the
pathology measures are very low and indeed much lower than others in her cohort. She has
quite strongly held beliefs in world benevolence and world meaningfulness with respect to
her cohort as well. The only negative things that can be gleaned from the comparison of
Geena’s scores to those of the other grandchildren in the study sample is that she scores
higher on the usage of maladaptive coping strategies and also higher on the negative
attachment dimension of attachment anxiety. However, despite her higher score on the
negative attachment dimension she can still be classified with a secure attachment style, as
her scores for the positive attachment dimensions are also quite high.
© Janine Lurie-Beck 2007
323
Table 15.16.
Geena’s scores compared to whole grandchild of survivor sample and relevant grandchild of survivor subgroups means
on psychological impact and influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
AAS Type
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Geena
All grandchildren
All with one
survivor
grandparent
Minimum
Score
Maximum
Score
Highest
Score in
Normal
Range
0.00 (normal)
2.00 (normal)
3.00
56.00
17.00
Secure
6.22
5.19
10.48
40.11
16.52
4.80
7.00
9.60
43.60
13.20
0.00
0.00
0.00
12.00
6.00
42.00
42.00
24.00
60.00
30.00
7.00
9.00
56.00
109.00
47.00
44.00
40.66
90.96
34.55
37.46
43.60
83.40
36.60
31.00
24.00
36.00
8.00
12.00
80.00
144.00
48.00
72.00
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
Table 15.17 below presents Geena’s perceptions of her child of survivor father compared to
her perceptions of her mother who is not descended from Holocaust survivors. As can been
seen, Geena rates both of her parents very favourably and there is little difference in the
ratings she gives for her father and mother. The ratings Geena gave her parents are also
more favourable than comparable sections of the grandchild of survivor sample. Table
15.18 presents Geena’s perception of cohesion and expressiveness/level of communication
within her family. She perceives a very high level of cohesion within her family; in fact
she obtains the highest possible score on this scale. Her rating of family expressiveness is
similar to ratings given by her contemporaries.
Table 15.17.
Geena’s perceptions of her child of survivor father versus her perceptions of her non-child of survivor mother
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
Geena’s
perception of
her child of
survivor
father
15.00
0.00
0.00
57.00
Grandchildren’s
perceptions of
child of survivor
fathers
Geena’s
perception of
her non-child of
survivor mother
10.44
2.88
5.63
50.41
14.00
0.00
0.00
57.00
Grandchildren’s
perceptions of
non-child of
survivor
mothers
13.00
0.83
1.33
58.33
Notes. PCS = Parental Caregiving Style Questionnaire, PAQ = Parental Attachment Questionnaire
Minimum
Score
Maximum
Score
0.00
0.00
0.00
14.00
16.00
16.00
16.00
70.00
Table 15.18.
Geena’s perceptions of family interaction patterns compared to whole grandchild of survivor sample and relevant
grandchild of survivor subgroups means
Geena
FES – Cohesion
FES – Expressiveness
68.00
54.00
Note. FES = Family Environment Scale
© Janine Lurie-Beck 2007
All grandchildren of
survivors
48.18
50.07
All grandchildren with a child of
survivor father only
62.00
52.40
Minimum
Score
1.00
15.00
Maximum
Score
68.00
73.00
324
15.3.2. – “Solange” - Grandchild with Four Survivor Grandparents
Solange was born in Australia in 1979 and was 26 years old when she completed the
questionnaires for this study. All four of her grandparents survived the Holocaust in some
way. They originated from Poland, Hungary, Romania and Lithuania. Her mother was
born seven years after the end of the war and her father was born four years after the war.
In her current life, Solange is single and has a tertiary education. She is not a member of
any descendant organisation. She reports that she has attended therapy in the past.
Solange’s scores on the psychological impact variables and influential
psychological processes are presented in Table 15.19. She scores within the normal range
for anxiety and depression, however in comparison to Geena (who has only one survivor
grandparent) she experiences more anxiety and depressive symptoms.
Her score on
vulnerability (which relates to her fear for the safety of herself and others from harm) is
relatively high and is slightly higher than her contemporaries and certainly also much
higher than Geena’s score of 3.00. Solange’s use of maladaptive coping strategies is higher
than the comparison grandchild of survivor groups, while her usage of adaptive coping
strategies is markedly lower. Solange is classified with a dismissing attachment style.
While it is of course possible that the differences between Solange and Geena are due to
factors other than the number of survivor grandparents they have it is hard to ignore the
confirmatory pattern of results.
Table 15.19.
Solange’s scores compared to whole grandchild of survivor sample and relevant grandchild of survivor subgroups means
on psychological impact and influential psychological process variables
Impact Variables
DASS Anxiety
DASS Depression
PTV
AAS Positive Dimensions
AAS Negative Dimensions
AAS Type
Influential Psychological Processes
COPE Maladaptive
COPE Adaptive
WAS - Benevolence
WAS - Meaningfulness
Solange
All Grandchildren
All with four
survivor
grandparents
Minimum
Score
Maximum
Score
Highest
Score in
Normal
Range
3.00 (normal)
5.00 (normal)
11.00
32.00
15.50
Dismissing
6.22
5.19
10.48
40.11
16.52
5.10
6.80
10.75
35.10
15.35
0.00
0.00
0.00
12.00
6.00
42.00
42.00
24.00
60.00
30.00
7.00
9.00
44.50
85.00
37.50
39.50
40.66
90.96
34.55
37.46
37.65
97.90
32.25
34.25
24.00
36.00
8.00
12.00
80.00
144.00
48.00
72.00
Notes. DASS = Depression Anxiety Stress Scales, PTV = Posttraumatic Vulnerability Scale, AAS = Adult Attachment
Scale, WAS = World Assumptions Scale
Table 15.20 provides a comparison of Solange’s perceptions of her child of survivor
parents both to each other and the perceptions of other grandchildren in the sample. While
© Janine Lurie-Beck 2007
325
she credits both her parents with a certain degree of warmth, she rates both of them quite
highly on ambivalence (which no doubt plays at least a partial role in her classification with
a dismissing attachment style). She rates her mother as notably less warm than her father.
Her father is rated warmer than other child of survivor fathers rated by the study sample,
but her mother is rated less warm than other child of survivor mothers. However, both
parents are rated relatively favourably (and equivalently to perceptions of her cohort) in
terms of their perceived encouragement of her autonomy and independence. Table 15.21
compares Solange’s perceptions of family cohesion and expressiveness to relevant
grandchild of survivor subgroups. Solange perceives a very low level of cohesion within
her family and she also rates family expressiveness lower than grandchild of survivor
subgroups. The overall portrait of her family life painted by Solange’s scores on the family
interaction variables is one of disengagement.
Table 15.20.
Solange’s perceptions of her child of survivor father versus her perceptions of her child of survivor mother
PCS – Warmth
PCS – Coldness
PCS – Ambivalence
PAQ – Fostering of Autonomy
Solange’s
perception of
her child of
survivor
father
13.00
3.00
10.00
51.50
Grandchildren’s
perceptions of
child of survivor
fathers
Solange’s
perception of
her child of
survivor mother
Grandchildren’s
perceptions of
child of survivor
mothers
Minimum
Score
Maximum
Score
10.44
2.88
5.63
50.41
8.50
3.00
10.00
48.50
11.89
1.41
3.86
48.86
0.00
0.00
0.00
14.00
16.00
16.00
16.00
70.00
Notes. PCS = Parental Caregiving Style Questionnaire, PAQ = Parental Attachment Questionnaire
Table 15.21.
Solange’s perceptions of family interaction patterns compared to the whole grandchild of survivor sample and relevant
grandchild of survivor subgroups means
Solange
FES – Cohesion
FES – Expressiveness
16.00
41.00
Notes. FES = Family Environment Scale
All grandchildren of
survivors
48.18
50.07
All grandchildren with two child of
survivor parents only
36.73
52.40
Minimum
Score
1.00
15.00
Maximum
Score
68.00
73.00
15.3.3. – Summary and Conclusions from Grandchild of Survivor Case Studies
The cases of Geena and Solange serve to emphasise how the influence of the Holocaust has
far from dissipated two generations on. While Geena scores very well on psychological
measures and also rates her parents favourably in terms of parent-child attachment
dimensions and their encouragement of her independence, she is still sufficiently influenced
by her family history to have joined a descendant organisation. The influence of that one
survivor grandparent and the knowledge of her experiences has certainly shaped her
identity to some degree.
© Janine Lurie-Beck 2007
326
The case of Solange highlights the probable effects of the Holocaust on family
dynamics and how that influences the children within these contexts. Solange certainly
rates her family environment quite negatively and her classification as having a dismissing
attachment style is certainly predictable from the environment she intimates through her
ratings. It is also interesting to note her relatively high score on vulnerability as well. She
notes that she often thinks about the war and perhaps this focus has contributed to her
heightened sense of self and family vulnerability.
15.4. – Summary and Conclusions
The case studies presented in this chapter underscore the wide range in the degree of postwar adaptation among survivors as well as the flow on effects to future generations. The
demographic variables of nature of Holocaust experiences, loss of family and country of
origin of the survivors themselves which were the highest ranking demographic variables in
terms of influence on psychological health (as identified in Chapters Thirteen and Fourteen
in the empirical study) were chosen as guidelines for the selection of case studies. Of
course, it is also acknowledged that the scores of the case study participants cannot be
solely due to Holocaust experience or ancestry but overall they (along with the quantitative
analysis of the study data) do relate to Holocaust-related variables in a predictable way and
so its potentially strong influence cannot be ignored.
In reading the details of the survivor and survivor ancestor’s Holocaust experiences,
it can be more fully understood how individual survivors may have endured quite different
events and traumas. However, what unites them all is that their experiences during the war
were as a result of their being targeted by the Third Reich, whether they succumbed to the
fate planned for them (camp incarceration) or escaped it (by living in hiding or escaping
Europe). With a diverse range of experiences, it is not unpredictable that there should be a
diverse range and severity of responses.
What the research in the current thesis has
demonstrated is the heterogeneity of the impact of the Holocaust, not only on survivors but
on their descendants as well. It is the reasons behind this heterogeneity that should have
been the focus of research from the outset, not determined attempts to collectively “pigeon
hole” the entire population of survivors and descendants as if they would display exactly
the same level of symptomatology and interpersonal difficulties.
© Janine Lurie-Beck 2007
327
Chapter Sixteen – Discussion and Conclusions
Over the course of the current thesis, the reader has been presented with large amounts of
data and analysis relating to a plethora of variables and how they relate (or not) to the
psychological health of Holocaust survivors and their descendants. What has been revealed
is a complex web of inter-related variables which perhaps may seem to uncover more
questions than answers.
Clearly the impact of the Holocaust on survivors and their
descendants is not readily defined by black and white boundaries; it is indeed tempered by
innumerate shades of grey. The Holocaust survivor population (and that of its descendants)
is definitely far from a uniform, homogenous group in terms of its post-war psychological
health and functioning. What is also clear is that the Holocaust had a profound influence
on the way survivors interacted with their children and that it continues to reverberate
through the generations. The extent to which this has occurred also differs widely within
the survivor descendant population.
The research presented in the current thesis has
attempted to delineate the variables that may help to explain this differential level of posttraumatic symptomatology and transmission of traumatic impact within the Holocaust
survivor and descendant population.
16.1. – Unique Contribution to the Holocaust Trauma Literature by the Current
Thesis
The body of research presented in the current thesis represents numerous unique
contributions to the literature with regard to the lasting effects of the Holocaust on
survivors and their descendants. These contributions include:
• the first meta-analysis of research comparing Holocaust survivors to the general
population (as represented by control groups);
• the first meta-analysis of research comparing grandchildren of survivors to the general
population;
• the first meta-analysis of research comparing children of survivors to the general
population to include both published and unpublished data.
• the first international study to attempt to follow the psychological impacts and potential
modes of trauma transmission across three generations
What is apparent is that overall survivors and their descendants do differ statistically
significantly from the general population, in terms of having a higher baseline level of
© Janine Lurie-Beck 2007
328
psychopathological symptoms. The higher baseline may well be still within the normal
range of symptom experience but it is measurable and discernible. However, what is also
apparent is that it is essentially meaningless to attempt to aggregate all survivors or all
descendents as if they are part of a homogeneous group – they clearly are not.
To this end, the current thesis has made numerous contributions in the assessment
of demographic differences within the survivor and descendant populations.
These
contributions include:
• the first meta-analytic synthesis of studies considering demographic differences within
the survivor and descendant populations;
• meta-regressions of existing data with respect to the influence of gender, age and time
lapse since the war among survivors which have provided new clarifications as to the
role of these variables;
• the first empirical examination of a number of demographic variables within the
survivor population such as country of origin;
• the first examination of the possible influence of post-war delay in the birth of children
of survivors and those children’s psychological health – examined both indirectly via
meta-regressions and directly with raw data in the empirical study.
• the first comparison of perceptions of survivor and non-survivor parents within one
survivor parent families and the first comparison of perceptions of child-of-survivor
parents and non-child-of-survivor parents within one child-of-survivor parent families.
The aggregation of data relating to numerous demographic factors has served to
both clarify the role of some, while highlighting the lack of knowledge that exists in
relation to many others because of widely differing operationalisation, methodological
flaws or lack of assessment.
16.2. – Thought-provoking Findings Emerging from the Current Thesis
A large number of analyses were presented in this thesis and to attempt to address each and
every one here would make this chapter as long as the rest of the thesis. In this section, a
number of interesting themes that have emerged throughout the course of the meta-analyses
and empirical study are articulated.
16.2.1. – The Role of Gender
The influence of gender (both individual and parental) on survivor and descendant
psychological functioning is a complex one. While gender differences among survivors
© Janine Lurie-Beck 2007
329
and descendants are relatively uniform in suggesting greater female vulnerability in terms
of their own psychological health, the role of gender within the parent-child relationship is
much more complicated.
According to the results of the empirical study reported in this thesis, the nature of
the mother-child relationship (in terms of warmth, coldness or ambivalence and
encouragement of the child’s independence) generally have a stronger influence on the
child’s psychological health than the father-child relationship.
However, greater
differentiation in child of survivors’ scores on psychological impact variables and
perceptions of survivor parents was based on the nature of survivor father’s Holocaust
experiences, rather than survivor’s mother’s experiences. In other words, while maternal
dimensions are more strongly related to childrens’ adult psychological profile and maternal
Holocaust experiences do create some differences in their care-giving behaviours, within
the lower baseline of paternal influence there is greater variation based on paternal
Holocaust experiences. This suggests that Holocaust experiences may well have a stronger
effect on female survivors in terms of mental health symptoms, but may have a stronger
effect on the male survivors’ parenting abilities and approaches which in turn translates to
greater variability in the transmission of traumatic impact to the children.
In addition, results of the empirical study suggest an interaction between survivor
parent and child of survivor gender in relation to the transmission of Holocaust trauma
cross-generationally. In other words, the degree to which negative effects of the Holocaust
infuse the parent-child relationship and the degree to which trauma transmission occurs
depends on the parent-child gender combination.
To reiterate, the strength of these
processes differ between mothers and daughters, mothers and sons, fathers and daughters
and fathers and sons. Specific details of which pairing is the most detrimental could not be
derived based on the relatively small empirical study data set, however this is certainly an
area worth exploring in more detail in further research.
16.2.2. – Country of Origin
Survivors’ country of origin was found to be the second most influential demographic
among the survivor generation and the most influential demographic among the children of
survivor generation. This variable has been relatively understudied within the Holocaust
literature, with only one study being found to consider its influence among survivors (using
© Janine Lurie-Beck 2007
330
only a Western versus Eastern Europe dichotomy) and only two studies located to consider
its potential influence on children of survivors.
The reasons why survivor country of origin was found to be so influential within the
empirical study conducted for this thesis are numerous. A survivor’s country of origin
determined the course of the events during the Holocaust as well as the duration of
persecution they experienced. Survivors from Poland had quite different experiences to
those from Hungary, who in turn had different experiences to those from the Netherlands.
It can be argued that the duration of persecution can be positively or inversely related to
post-Holocaust symptom levels, depending on whether longer more gradual periods of
persecution are seen as more traumatising, or provide a process of acclimatisation not
afforded to those who experienced a shorter and more rapid persecution timeline.
Quite apart from the differing Holocaust experiences, are the cultural differences
that exist (and existed prior to the Holocaust) between people from different countries, of
different nationalities, racial and cultural groups. What proportion of the variation found
that is based on country of origin is really attributable to cultural differences as opposed to
experiences during the Holocaust? Are there differences in resiliency between people from
different cultures, perhaps attributable to some form of national character? Are differences
in attachment dimensions reflective of existing cultural differences with the comfort with
closeness and dependence on others, the need for companionship, the desire for
independence, the nature of parent-child relationships?
The extent to which a survivor felt able to return to their home country after the war
also differed by country. For example, Jewish survivors wanting to return home were
greeted with far from welcoming arms in some Eastern European countries such as Poland,
while those from countries such as the Netherlands or France may have felt more
comfortable and less vulnerable. Therefore, could it be that the reception they received
upon their attempted home-coming that further compounded their symptomatology as
opposed to the Holocaust traumas themselves? Certainly this was found to be the case for
returning Vietnam Veterans (de Silva, 1999; Lomranz, 1995; McCann & Pearlman, 1990;
Z. Solomon, 1995).
It was not possible to delineate the proportion of influence attributable to culture,
differential Holocaust timelines or post-Holocaust reception within the empirical study in
the current thesis. What is clear though, is that discernible differences in post-Holocaust
© Janine Lurie-Beck 2007
331
symptom levels can be ascribed to a survivor’s country of origin, whatever the specific
source.
It would be desirable, but not possible, to compare means obtained in the
empirical study to normative data for each European country relevant. Unfortunately,
separate normative data for different European countries does not exist for the measures
used in the current thesis.
It is certainly acknowledged, however, that the relative
importance of cultural differences and the role of the Holocaust on symptom measures and
family interaction variables is something worthy of investigation.
16.2.3. – The Impact of Post-war Delay in Child-rearing
The use of meta-regression techniques allowed the assessment of demographic variables
that to date had not been tested including the post-war delay in birth of children of
survivors. Time lapse between the end of the war and the birth of the second generation
has been an issue often mentioned within theoretical conjecture and anecdotal discussions
of clinical experiences. The essence of arguments about its hypothesised influence relates
to the degree to which survivors had recovered from, or processed, their traumatic
experiences during the Holocaust, as well as the extent to which they suffered from
unresolved mourning. A negative relationship between the post-war delay in birth and the
extent to which children of survivors suffer from symptomatology was proposed but never
assessed directly.
Within this thesis, the issue of post-war delay in birth was assessed indirectly via
meta-regression, as well as directly within an empirical study. While results from the metaregressions largely supported the assertion that shorter delays in birth are associated with
higher levels of symptoms among children of survivors, the results from the empirical
study, where significant, tended to suggest the opposite.
However, the effects of post-war delay in birth of children of survivors also related
to their children’s (that is, the grandchildren of survivors) perceptions of them as parents.
Specifically, sons of survivors who were born after a longer delay were viewed as more
encouraging of their children’s independence, and family expressiveness and cohesion were
also inversely related to delay in birth. Therefore the delay in the birth of children of
survivors may well have a bearing not only on their level of symptomatology, but also on
the way in which they interact and relate with their children.
Analysis of the influence of post-trauma delay in birth of children of trauma
survivors is by no means definitive if it is based solely on the results of the meta-analyses
© Janine Lurie-Beck 2007
332
and empirical study reported in this thesis. The findings reported in this thesis, on this
issue, certainly highlight the need for survivors/victims of trauma to be encouraged to allow
themselves time to process and recuperate from their ordeal before embarking on the
already stressful role of parenthood.
16.2.4. – The Compounding Traumatic Impact among Survivor Dyads
An interesting finding is that among two survivor or child-of-survivor parent families, both
survivor or child-of-survivor mothers and survivor or child-of-fathers are rated as colder,
more ambivalent and less encouraging of independence/autonomy than when the other
parent is not a survivor or child-of-survivors. In other words, while survivor or child of
survivor parents are viewed more negatively than non-survivor or non-child-of-survivor
parents, in general, the effect of the other parent also being a survivor or descended from
survivors further compounds the negative perceptions of the parents.
Of course, it cannot be known for sure whether it is the children’s perceptions that
are being affected or whether survivor or child-of-survivor-parents truly are colder, more
ambivalent and less encouraging of independence if they are part of a survivor or child-ofsurvivor couple. What is interesting is the idea that each individual within the family
system has the ability to not only affect the system as a whole but potentially other
individuals within the system.
Truly no survivor or descendant of survivors can be
understood within an individual vacuum – they must be considered within their family
historical context.
16.2.5.
–
The
Influence
of
Post-war
Settlement
Location
on
post-Holocaust
Symptomatology
After the end of the war, Holocaust survivors settled all over the world. Some remained in
Europe, but others moved to far away continents such as America or Australia, while others
moved to Israel. Differences in post-war adjustment related to post-war settlement location
have barely been examined in the literature.
This omission is startling given the
acknowledgement by many in the trauma field that the recovery environment of trauma
survivor can be very important in determining how well they recover (B. L. Green et al.,
1985; Wilson, 1989).
In the current thesis, the influence of post-war settlement location on survivor’s
psychological health was examined indirectly via sub-set meta-analyses (comparing groups
of studies based on the country in which they were conducted) as well as directly via the
© Janine Lurie-Beck 2007
333
empirical study. While the sub-set meta-analyses largely afforded comparisons between
survivors who settled in America to those who settled in Israel, the empirical study
presented the first comparison of survivors who settled in America to those who settled in
Australia or New Zealand. Within the empirical study it was the survivors who settled in
Australia or New Zealand who reported higher depression and anxiety and statistically
significantly higher levels of PTSD symptoms than those who settled in America.
The first conclusion may well be to suggest that survivors who settled in Australia
or New Zealand were not received as well as they were in America. However, another
possibility exists. Could it be that the survivors (or at least the participants in the empirical
study) who settled in Australia were the more traumatised to begin with? Survivors often
talk about wanting to get as far away from Europe as possible. Is the distance between
Europe and a survivors’ choice of post-war settlement location directly related to the
severity of their traumatic experiences and/or the severity of symptomatology they
experienced as a result?
16.2.6. – The Case of the Grandchildren of Holocaust Survivors
Grandchildren of Holocaust survivors are only now just reaching adulthood in large enough
numbers to enable relatively large-scale studies of their psychological health. A lay person
might be forgiven for assuming that, two generations on, the grandchildren of survivors
should be no different to other members of the general population. However the results
from both the meta-analyses and the empirical study conducted for this thesis suggest that
the legacy of the Holocaust continues to be felt.
The fact that both meta-analytic results, as well as results of the empirical study
conducted for the current thesis, suggest an upturn in symptom prevalence and severity for
this group is a worrying finding. Is this finding reflective of a pattern in the wider
community, or does it belie a resurgence of the impact of ancestral trauma? Certainly with
each generational separation from the Holocaust, there are more and more intervening nonHolocaust related traumas and events that could impinge on the incidence of symptoms and
this must be borne in mind. However, this upturn was not only noted in the empirical study
when comparing grandchildren to the previous generations. It was also seen in the metaanalyses which compared them to members of their own generation who do not have
survivor ancestors and found greater disparity between grandchildren of survivors and their
contemporaries than between children of survivors and their contemporaries. This suggests
© Janine Lurie-Beck 2007
334
that this upturn in symptoms is not merely reflecting a community-wide generational
pattern, but possibly a genuine increase in symptoms related to survivor ancestry.
One possible explanation is that grandchildren of survivors are disproportionately
more despairing about the current world climate within the context of the knowledge of the
suffering of their own family members. Does the knowledge of the Holocaust, or more
specifically the knowledge of how individuals were affected by it, impel the grandchild of
survivors to greater anxiety about its repetition or repetition of like-suffering around the
world?
Krasnostein (2006) conducted a number of workshops with grandchildren of
survivors and found that issues pertaining to the Holocaust still resonate profoundly for this
generation. She notes that grandchildren in the workshop groups commented on many
difficulties associated with their identity as a grandchild of Holocaust survivors such as
“the need for inner healing and family healing, dealing with anger, wanting to right the
wrongs, denying effects, dealing with Holocaust denial, feeling burdened with the
responsibility of carrying on the legacy, pondering notions of control, guilt, pressure to
achieve and of course dealing with the good old fashioned persecution complex.” Many
think about “existential concepts such as If it weren’t for Hitler I wouldn’t be me as Hitler
was my grandparent’s matchmaker (Krasnostein, 2006).”
16.3. – Revised Model of the Differential Impact of Holocaust Trauma across Three
Generations
The Model of the Differential Impact of Holocaust Trauma across Three Generations was
formulated and tested via meta-analysis and empirical study over the course of the current
thesis. Table 16.1 reports on the hypotheses regarding the role of influential psychological
processes and family interaction processes/potential modes of trauma transmission. As can
be seen, these hypotheses were generally supported by the results of the empirical study.
© Janine Lurie-Beck 2007
335
Table 16.1.
Status of Hypotheses relating to relationships between model variables
Model Hypotheses
Empirical Study Findings
MH1: Negative/dysfunctional coping strategies will be
This hypothesis was supported. The 14 coping strategies
positively related to negative psychological symptoms and
measured by the COPE were combined into maladaptive
negatively related to positive psychological dimensions,
coping and adaptive coping composites. Maladaptive
while positive/functional coping strategies will be negatively
coping strategies were found to be stronger predictors of
related to negative psychological symptoms and positively
scores on psychological impact variables than adaptive
related to positive psychological dimensions.
coping strategies.
MH2: Strength of belief that the world is benevolent and
This hypothesis was supported. A belief in world
meaningful will be negatively related to negative
benevolence (that the world is a kind and caring place)
psychological symptoms and positively related to positive
was more strongly related to psychological impact
psychological dimensions.
variables than world meaningfulness (that the world is an
understandable and predictable place).
MH3: Posttraumatic growth aspects will co-exist with
This hypothesis was supported in relation to scores on the
negative psychological symptoms (in other words
IES-R which measures PTSD symptoms; however there
posttraumatic growth and negative psychological symptoms was a negative correlation with depression, particularly
will be positively related).
with the personal strength growth aspect.
MH4: Negative parent-child attachment dimensions such as Parental warmth was uniformly related to decreases in
the degree of coldness and ambivalence will be positively
negative symptomatology while parental coldness and
associated with negative psychological symptoms and
ambivalence were uniformly related to increases in
negatively associated with positive psychological symptoms symptom levels. There was a much stronger relationship
and positive parent-child attachment dimensions such as
with maternal dimensions as opposed to paternal
perceived parental warmth will be negatively associated with dimensions.
negative psychological symptoms and positively associated
with positive psychological dimensions.
MH5: A curvilinear/U-shaped relationship will exist between
U-shaped relationships were noted for depression and
negative psychological symptoms and family cohesion (with anxiety among children of survivors, however linear
very low and very high cohesion associated with higher
relationships were noted for negative attachment
symptom levels than mid-range scores) and an inverted Udimensions and adaptive coping strategies (negative
shaped relationship will exist between positive psychological linear relationships) and positive attachment dimensions
dimensions and family cohesion.
and belief in world benevolence (positive linear
relationships).
MH6: The degree to which parents are encouraging of their Maternal encouragement of independence was negatively
children’s attempts to establish independence will be
related to depression and negative attachment
negatively associated with negative psychological
dimensions and positively related to positive attachment
symptoms and positively associated with positive
dimensions and belief in world benevolence. This
psychological dimensions.
parenting dimension was not as strongly related to
children’s scores as the parental attachment dimensions
of warmth, coldness and ambivalence.
MH7: General communicativeness within the family unit will This hypothesis was supported although the strength of
be negatively associated with negative psychological
these relationships was not as strong as for other
symptoms and positively associated with positive
variables.
psychological dimensions.
MH8: Negative modes of communicating about the
The communication modes of affective communication
Holocaust, such as guilt-inducing, indirect and non-verbal
(particularly maternal) and non-verbal communication
will be positively associated with negative psychological
about the Holocaust were the most strongly related to
symptoms and negatively associated with positive
children of survivor’s scores on psychological impact
psychological symptoms and positive modes of
variables. As predicted these negative modes of
communicating about the Holocaust, such as frequent,
communication were associated with increased negative
willing and open discussion will be negatively associated
symptoms and decreased scores on positive variables.
with negative psychological symptoms and positively
associated with positive psychological dimensions.
Mediation of the relationship between parent and child
Unable to test the mediation due to inadequate sample
scores on psychological impact variables by parent-child
numbers.
attachment, family cohesion, parental encouragement of
child’s independence, family communication and
communication about the Holocaust.
© Janine Lurie-Beck 2007
336
Figure 16.1 presents the revised version of the model as it stands at the conclusion
of the assessments conducted in the current thesis. A rank ordering of the importance of
influential psychological processes, proposed transmission modes and demographic
variables in predicting the psychological impact of the Holocaust on survivors and
descendants was achieved. It is acknowledged that other processes may also play a role in
determining the psychological health of survivors and their descendants. Certainly there
are other dimensions of family environment which could be measured, as well as other
influential cognitive processes. An exhaustive list of such variables was not included in
this thesis. The main focus was on exploring the demographic variables that potentially
moderate not only the impact of the Holocaust on survivors, but also the way they
interacted with their children and therefore the transmission of trauma across generational
boundaries. While the current thesis examined many demographic variables (a number of
which had either been inadequately assessed or not assessed at all in the existing literature)
there are a number of variables that could not be analysed due to lack of data or small
sample sizes. The demographic variables mentioned within the anecdotal literature that
could not be analysed have been noted in the version of the model in Figure 16.1.
© Janine Lurie-Beck 2007
337
Psychological
Impacts of the
Holocaust
Depression
Anxiety
Paranoia
PTSD symptoms
Romantic
Attachment
Dimensions
• Post-traumatic
Growth
3rd Generation
(Grand-children of Survivors)
2nd Generation
(Children of Survivors)
1st Generation
(Survivors)
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Depression
Anxiety
Paranoia
Romantic
Attachment
Dimensions
Depression
Anxiety
Paranoia
Romantic
Attachment
Dimensions
Ranking of Influential
Psychological Processes
Ranking of Modes of
Intergenerational Transmission of
Trauma
Ranking of Demographic Moderators
Holocaust
Survivor Generation
Children of Survivor
Generation
Grandchildren of
Survivor Generation
(1) Nature of experiences
(2) Country of origin
(3) Loss of family during the Holocaust
(4) Post-war settlement location
(5) Age during the Holocaust (in 1945)
(6) Gender
(7) Length of time before resettlement
(1) Maladaptive coping strategies
(2) Assumption of World
Benevolence
(3) Assumption of World
Meaningfulness
(4) Adaptive coping Strategies
Unable to test:
Reason for persecution
Time lapse
(1) Maladaptive coping strategies
(2) Assumption of World
Benevolence
(3) Assumption of World
Meaningfulness
(4) Adaptive coping Strategies
(1) Maladaptive coping strategies
(2) Assumption of World
Benevolence
(3) Assumption of World
Meaningfulness
(4) Adaptive coping Strategies
(1) Parent-Child Attachment
(especially maternal)
(2) Family Cohesion
(3) Communication about Holocaust
experiences (specifically via
affective or non-verbal modes)
(4) Encouragement of
Independence (maternal)
(5) General Family Communication
(1) Parent-Child Attachment
(especially maternal)
(2) Family Cohesion
(3) Encouragement of
Independence (maternal)
(4) General Family Communication
(1) Parent country of origin
(2) Parent nature of Holocaust experiences
(3) Parent loss of family during the Holocaust
(4) Survivor Parent gender
(5) Parent post-war settlement location
(7) Parent age during the Holocaust (in 1945)
(8) Number of survivor parents
(6) Post-war delay in birth
(9) Birth order
(10) Birth before or after
parent settlement outside of
Europe
(11) Gender
Unable to test:
Reason for persecution of parent/s
(3) Number of survivor grandparents
(4) Grandparent post-war settlement location
Unable to test:
Grandparent age during the Holocaust
Grandparent gender
Grandparent type/nature of Holocaust experiences
Grandparent loss of family
Grandparent country of origin
Length of time before grandparent
resettlement/time spent by grandparent in displaced
persons camps
Reason for persecution of grandparent/s
(1) Number of child of survivor
parents
(6) Delay between the end of
the war and the birth of
parents
(5) Parent gender
(7) Birth order
(2) Gender
Unable to test:
Birth of parent/s before or
after survivor grandparent/s
emigration
Parent birth order
Figure 16.1. Revised Model of the Differential Impact of Holocaust Trauma across Three Generations
© Janine Lurie-Beck 2007
338
One of the initial stated aims of this thesis was to recognise the most vulnerable
subgroups of the survivor and descendant populations. While Figure 16.1 denotes the
importance of the demographic factors, it does not delineate which demographic groups are
more affected than others. Tables 16.2 to 16.4 provide a visual representation of the most
vulnerable and most resilient subgroups relating to the demographic variables examined in
this thesis.
Table 16.2
Delineation of most and least affected demographic subgroups of survivors
Demographic Variable
Least affected Subgroups
Holocaust experiences
In hiding, escaped, partisan
Gender
Males
Survival of other family members
Some surviving family members
Experience of Holocaust alone or
Always with at least one family
with family
member
Post-war settlement location
America
Age
Younger survivors
Country of origin
Belgium, Netherlands
Most affected Subgroups
Concentration/Labour Camp
Females
Sole survivor
Spent some time without any family
members
Australia
Older survivors
Hungary, Poland, Lithuania, Ukraine
Table 16.3
Delineation of most and least affected demographic subgroups of children of survivors
Demographic Variable
Least affected Subgroups
Most affected subgroups
Gender
Males
Females
Number of survivor parents
One
Two
Gender of survivor parent
Mother
Father
Post-war delay in birth
Ambiguous results
Birth order
Middle order
Survivor parent country of origin
Belgium, Netherlands
Hungary, Poland, Lithuania, Ukraine
Survivor parent survival of family
Some surviving family members
Sole survivor
members
Survivor parent nature of Holocaust
Non-camp
Camp
experiences
Table 16.4
Delineation of most and least affected demographic subgroups of grandchildren of survivors
Demographic Variable
Least affected Subgroups
Most affected subgroups
Number of survivor grandparents
One survivor grandparent
Four survivor grandparents
Number of child of survivor parents
One
Two
Gender
Males
Females
Birth order
Middle order
16.4. – Applicability and Adequacy of Existing Trauma Theory in Explaining PostHolocaust Adaptation among Survivors
Green, Wilson and Lindy’s (1985) Working Model for the Processing of a Traumatic Event
and Wilson’s (1989) Person-Environment Interaction Theory of Traumatic Stress Reactions
were referred to in the Model of the Differential Impact of the Holocaust across Three
Generations which was developed and refined through the course of the current thesis. In
© Janine Lurie-Beck 2007
339
particular, these two theories were used as a basis to explain differences in post-Holocaust
adjustment among the survivor population: the people who were directly traumatised by the
Holocaust.
Green et al. (1985) and Wilson (1989) listed numerous elements/dimensions of a
traumatic experience that may explain differences in the extent to which the trauma
victim/survivor suffers from negative psychological symptoms after the traumatic event/s.
Green et al.’s (1985) model referred to aspects of the traumatic experience, such as the
degree of bereavement and the role of the survivor. In addition, they included elements of
the “recovery environment” such as intactness of community, societal attitudes, and
cultural characteristics. Wilson’s theory (1989) included a much more detailed list of
“environmental and situational variables” from which the demographic variables, included
in the Model of the Differential Impact of the Holocaust across Three Generations, were
additionally adapted.
Many of the dimensions listed in both models/theories were
applicable to all Holocaust survivors (such as life threat, exposure to death/dying, whether
the trauma was a community-based/collective trauma, whether the trauma was natural or
“man-made”, and whether the trauma consisted of single or multiple stressors) and so were
not included in the Model of the Differential Impact of the Holocaust across Three
Generations developed in the current thesis.
The demographic/situational variables representing specific elements of a traumatic
experience (presented in Green et al.’s (1985) and/or Wilson’s (1989) theories) that could
be used to differentiate subgroups of Holocaust survivors were included in the Model of the
Differential Impact of the Holocaust across Three Generations developed in the current
thesis. Each of these dimensions that could be applied to the Holocaust survivor population
will be discussed in turn. These dimensions are presented in Table 16.5
Table 16.5.
Elements from Green et al.’s (1985) and/or Wilson’s (1989) theories applied to the model of Holocaust trauma developed
in the current thesis
Green et al (1985)
Wilson (1989)
Included in current thesis model as…
Degree of bereavement
Degree of bereavement
Sole survivor status
Alone or with others
Time spent without family members
Role in trauma
Nature of Holocaust experiences
Intactness of community, Societal
Impact on community, Cultural rituals Country of origin
attitudes
for recovery, Societal attitudes
Both Green et al. (1985) and Wilson (1989) cited the degree of bereavement as a
consequence of the traumatic experience as being related to post-traumatic adjustment.
© Janine Lurie-Beck 2007
340
Degree of bereavement within a traumatic experience was certainly found to be of high
importance in determining post-traumatic symptomatology within the Holocaust survivor
population in this PhD research. In addition to bereavement, Wilson (1989) cites whether
the trauma victim/survivor was alone or with others during their traumatic experience as an
important variable. By definition, Holocaust survivors experienced the Holocaust with
others and not alone; however, within the current thesis this concept was narrowly and
specifically measured in terms of whether the survivor was alone for any part of their
wartime experiences or whether they were always with other family members.
This
variable, along with the related bereavement issue (as measured in the current thesis’
empirical study as whether the survivor was the sole survivor of their family) proved to be
an important determinant of post-Holocaust well-being.
Green et al. (1985) and Wilson (1989) also referred to a trauma victim/survivors’
role during the traumatic experience as being important. This has been borne out in
research with other populations, with respect to the benefits of an active versus a passive
role among displaced civilians versus political prisoners in the former Soviet state of
Georgia (Makhashvili et al., 2005) and Lithuania (Kazlauskas et al., 2005), survivors of
industrial accidents (Weisaeth, 2005) and bushfires (Parslow, 2005). In the one study
located that considered this issue among Holocaust survivors, Favaro et al. (1999) found
compatible results (see Chapter Eight, Section 8.2). It was not possible to directly measure
this concept within the empirical study presented in the current thesis because all the camp
survivors, who participated in the study, were interned for ethnic/religious reasons,
therefore providing no sub-group interned for active resistance of the Nazi regime. It is
regrettable that this issue could not be explored further within the Holocaust survivor
sample in the current thesis. Future researchers are encouraged to follow up on this
important dimension of traumatic experience within the Holocaust survivor population.
In addition, Green et al. (1985) and Wilson (1989) included a number of
dimensions which are defined by/related to a survivor’s country of origin. Specifically
these variables are the location of the survivor during the trauma, the duration/severity of
the trauma and the impact on the community (Wilson, 1989), intactness of community, and
cultural characteristics (B. L. Green et al., 1985). As was discussed in Chapter Four,
Section 4.4, a survivor’s country of origin had implications for the number of years they
endured persecution, the speed with which persecution was enacted (or in other words the
© Janine Lurie-Beck 2007
341
number of intervening stages, particularly a gradual removal of civil liberties and
ghettoisation, before camp incarceration). The results pertaining to country of origin of
survivors in the empirical study presented in the current thesis are quite compelling and
country of origin was ranked in the top three demographic variables in terms of the extent
to which they related to psychological health. The finding that survivors from Hungary
tended to generally present with higher symptomatology levels points to the speed of
persecution enactment being of more importance than the duration of the persecution.
Survivors from Hungary were hurriedly moved into ghettos and many by-passed these
altogether to be moved straight to concentration camps from 1944 onwards.
Survivors
from Eastern European countries endured a period of what might be labelled as restriction
of movement and living standards, followed by months or even years of ghettoisation
before being moved through the camp system. Wilson’s (1989) inclusion of the impact on
the community and what Green et al. (1985) label as the intactness of the community is also
to some extent measurable by the survivor’s country of origin, since survivors from
Western European countries overall were more able and willing to return to their
communities than those from Eastern European countries. Clearly, Holocaust survivor
country of origin and the highly related concept of speed of persecution enactment are
important determinants of post-war adjustment. It is believed that the empirical study in
the current thesis represents the first detailed assessment of the impact of country of origin
among Holocaust survivors, which given its revealed importance, leaves a gaping hole in
the Holocaust literature on a possibly extremely important factor.
Within the “post-trauma milieu” or “recovery environment”, Wilson (1989) lists
cultural rituals for recovery, while both Wilson (1989) and Green et al. (1985) cite societal
attitudes towards the event as important elements. Post-war settlement location was a
variable included in the current thesis’ Model of the Differential Impact of the Holocaust
across Three Generations which indirectly examines these elements. Certainly cultural
rituals for recovery were much more readily available to survivors in Israel than those who
settled in other countries. As was outlined in Chapter Four, Section 4.9.3, events such as
Yom Hoshoah provided survivors in Israel with a number of opportunities to participate in
collective mourning and recognition, and therefore receive validation of their suffering
during the war. Such opportunities were not available to the same extent to survivors who
settled elsewhere.
The meta-analytic results addressing post-war settlement location
© Janine Lurie-Beck 2007
342
(Chapter Eight, Section 8.8) were somewhat ambiguous with survey and incidence studies
suggesting opposing findings.
Unfortunately, it was not possible to compare survivors who settled in Israel to
those who settled elsewhere as part of the empirical study because only one survivor
participant was an Israeli resident. Interestingly, the one recent study to compare survivors
in Israel to those who settled in America (Kahana et al., 2005) found that survivors in Israel
scored uniformly lower than survivors in America on psychopathology measures. Given
that the number of society level rituals for recovery are much more prevalent in Israel, one
could hazard the suggestion that such aids have proved helpful; especially given the
continuing dangers of living in Israel (with countless wars and terrorism campaigns
between Israelis and Palestinians and surrounding Arab nations), one might predict that
survivors there would score higher on such measures than those who settled elsewhere.
Both Green et al. (1985) and Wilson (1989) also included coping strategies as
important determinants in post-trauma adjustment as well, and coping strategies were
included in the Model of the Differential Impact of the Holocaust across Three Generations
as influential psychological processes. As would be predicted by both models and a large
body of literature in the field, coping strategies were found to relate strongly to the
psychological health of survivors. In particular it was found that the use of maladaptive
coping strategies was much more strongly predictive of psychological symptom levels than
adaptive coping strategies were. The inclusion of coping strategies as an important variable
in predicting post-traumatic psychological health is vindicated for both Green et al. (1985)
and Wilson (1989) as well as the Model of the Differential Impact of the Holocaust across
Three Generations. Future research is encouraged to further delineate the specific coping
strategies that are the most predictive. The empirical study reported in the current thesis
collapsed the coping strategies measured into maladaptive and adaptive coping strategies
because of sample size restrictions.
Green et al.’s (1985) specification of post-traumatic growth as being diametrically
opposed to pathological outcomes, however, was not supported by the findings of the
current research, or indeed by numerous other studies within the literature (Cadell et al.,
2003; Laufer & Solomon, 2006; McGrath & Linley, 2006; Morris et al., 2005). Indeed,
posttraumatic growth was found to co-exist with negative symptomatology.
This
relationship does appear paradoxical on the surface and it is understandable that one might
© Janine Lurie-Beck 2007
343
theorise that posttraumatic growth and pathological outcomes are polar opposites. It seems
that if a traumatic event is sufficiently traumatising that it leads to chronic symptomatology,
it is also such a watershed moment in the victim/survivor’s life that it becomes a moment of
epiphany leading to sometimes dramatic changes in a person’s priorities.
16.5. – Contributions to Trauma Theory by the Research Presented in the Current Thesis
The inclusion of world assumptions as a potential influential psychological process in the
author’s Model of the Differential Impact of the Holocaust across Three Generations lead
to the uncovering of the important role this construct plays in affecting psychological
health. Indeed the strength of belief a person holds in the world as a kind and caring and
predictable and fair place were revealed to be more important in determining scores on
psychological impact variables than the use of adaptive coping strategies. This finding held
true for all three generations tested in the empirical study.
Experiencing a traumatic event/series of events as overwhelming as the Holocaust
inevitably must lead to some degree of shattering of world assumptions. The world was
proven to not be a kind and caring place and certainly the tenet that only bad things happen
to bad people and that when negative events befall a person they do so in a predictable way
could no longer be believed by a persecuted person during the Holocaust. Remarkably,
many of the survivors who participated in the empirical study presented in the current thesis
recorded relatively strong beliefs in world assumptions. It is not known whether these
survivors maintained their world assumptions during the Holocaust or have gradually built
them up again in the post-Holocaust years. Either way, this suggests somewhat of a
triumph of the human spirit. However, some sub-groups of the survivor population had
much weaker beliefs in the world assumptions and it was these groups who also suffer
more from negative symptomatology.
Clearly, the degree to which survivors' world
assumptions could be maintained or rebuilt are important determinants of the extent to
which they suffer from post-traumatic symptomatology. The transference/applicability of
this finding to other trauma populations is highly probable and is an area that warrants
further investigation.
While a number of the demographic and situational variables presented in Green,
Wilson and Lindy’s (1985) Working Model for the Processing of a Traumatic Event and
Wilson’s (1989) Person-Environment Interaction Theory of Traumatic Stress Reactions
could be applied to the Holocaust survivor population there are a number of more
© Janine Lurie-Beck 2007
344
Holocaust-specific dimensions that needed to be included in the Model of the Differential
Impact of the Holocaust across Three Generations.
The ranking of demographic/situational variables, in terms of their relationship to
psychological health and functioning among survivors, is something not presented by other
trauma models attempting to explain differential post-trauma adjustment. Other models
have tended to simply list such elements rather than determining which have more weight
in determining post-traumatic mental health. The ranking of the three elements intrinsic to
the Holocaust experience itself, namely nature of Holocaust experiences, loss of family and
country of origin point to the overarching importance of elements of the traumatic
experience. However, it is interesting to note that the post-war settlement location of
survivors was ranked as more important in explaining symptom levels than fundamental
demographic variables such as gender and age during the Holocaust. This suggests that the
post-traumatic environment and the differing receptions afforded to survivors of trauma can
be quite a powerful factor in explaining post-trauma adjustment.
It is important to acknowledge that a population of survivors of a large scale statebased traumatic event or period such as the Nazi Holocaust, or indeed more recent
analogous attempted genocides, will not react and recover in a uniform or homogenous way
and should not be treated as such. Sub-groups and experience subsets will be related to
differing levels of vulnerability and resilience. The key is to develop a framework to help
identify them so that help can be targeted and to be open to the idea that differences will
occur.
Within the Holocaust survivor group, particular vulnerability appears to lie with
survivors of camp internment (as opposed to survivors who were in hiding or had other
non-camp experiences), and the experience of at least part of the Holocaust without family
members.
That different traumatic events/sequences of events were experienced by
survivors of different nationalities and that post-war adjustment also differs by this was
hinted at, with data pointing to Hungarian survivors and their descendants evidencing the
highest symptom levels. However, given the small sample sizes these results are based on,
once groups were sub-divided by country of origin, it is not wise to make a definitive
statement on this variable until further data is available to confirm the findings presented in
the current thesis.
© Janine Lurie-Beck 2007
345
The vulnerability of highly bereaved and camp survivor subgroups translates into
vulnerability for their children in a relatively uniform way. This is perhaps not surprising
since these survivor subgroups are also rated more negatively by their children in terms of
family dynamics and communication about the Holocaust.
16.6. – Applicability of Attachment and Family Systems Theory in Understanding the
Intergenerational Transmission of Holocaust Trauma
The results of both the meta-analyses and the empirical study reported in the current thesis
certainly point to the necessity of appreciating the family background of descendants of
Holocaust survivors in understanding their presenting symptomatology. The basic tenet of
family systems theory (P. Minuchin, 1985) that an individual must be considered within the
context of their family system is highly relevant for descendants of survivors. This is
particularly pertinent when one is reminded that elements of survivor parent’s Holocaust
experience were the highest ranked demographic variables in explaining the symptom
levels of children of survivors, and the extent of survivor ancestry was also the highest
ranked demographic variable for the grandchildren of survivors. There has never been a
stronger example of how a traumatised person also affects the family system of which they
are a member – the negative fall out from which reverberates for generations.
The importance of the parent-child attachment relationship in influencing future
symptomatology levels among the grown-up children was also supported in the empirical
study conducted for the current thesis. Perceived parental (in particular maternal) warmth
was consistently related to a decrease in symptom levels, while parental (in particular
maternal) coldness and ambivalence were related to an increase in symptom levels. Indeed,
parent-child attachment dimensions were the most strongly related to the levels of
symptomatology experienced/reported by the grown-up children of survivors. Evidence for
the strong relationship between parent-child attachment and future symptom levels among
children within the general population has been consolidated within the literature, and the
findings of the empirical study in the current thesis further cement the base of proof.
A note should be included here that Wilson (1989) referred to an intensification of
developmental stages as being a consequence of experiencing a traumatic event. This is
akin to what many have argued is iconic of Holocaust survivor families – that the making
and breaking of ties within these families is particularly difficult (Kellerman, 2001b; Wardi,
1994). The separation-individuation stage (where young adult children move to assert their
© Janine Lurie-Beck 2007
346
independence from their family of origin culminating in their physical move from the
family home) was tacitly measured in the empirical study via the measurement of parental
encouragement of independence. While this variable was statistically significantly related
to a number of symptom variables, it was ranked third behind parent-child attachment
dimensions and family cohesion in terms of the strength of its relationship with
psychological impact variables among survivor descendants.
It should also be noted that the relationships between symptomatology and family
interaction variables (and the stronger influence of the maternal line) found among the
survivor population in the current thesis are not necessarily different from the relationships
between these variables in the normal population.
Chapter Three discussed the
relationships between family interaction variables and symptom levels among the general
population. The findings of the current thesis serve to highlight that traumas can impact on
family interaction variables (directly and indirectly) and that, when these are affected in a
negative way, they have the predictable negative effects on the children of that family.
Indeed, what we know about the relationships between family environment and mental
health makes the transmission of the impact of Holocaust or any other trauma highly
predictable.
16.7. – Key Role played by Communication about Holocaust Experiences
What is also apparent when examining the family interaction patterns within survivor
families is the key role played by communication about Holocaust experiences on the part
of survivors. Undoubtedly, experiencing the Holocaust led its traumatised survivors to
interact with their children in a different way (either more withdrawn or more clingy) than
they would have had they not gone through that trauma. However, open communication
about one’s experiences during the Holocaust helped survivors’ children understand why
their parents behaved and responded in the way they did and thus served to ameliorate the
detrimental effects of these negative family dynamics.
Communication about the Holocaust between survivors and their children plays a
vital part in determining the psychological health of the children. Indeed, communication
about the Holocaust was ranked as more important in determining children of survivors’
symptom levels than general communication within their family of origin within the
empirical study conducted for the current thesis. It is the extent and way in which their
survivor parents spoke to them about their Holocaust experiences which is more important
© Janine Lurie-Beck 2007
347
than their communication about anything else. Clearly, open discussion about Holocaust
experiences helps children of survivors to understand why their survivor parents behave
and react the way they do. A lack of ambiguity in these areas is helpful in ensuring that
children of survivors do not develop generalised objectless anxieties and helps them to
understand why their parents may be over-protective and clingy. Perhaps what is more
important is the way in which the Holocaust is discussed – particularly given that the use of
affect-laden communication was the most important of the Holocaust communication
modes measured in the empirical study.
Certainly children of survivors have a strong yearning to know and understand the
details of their survivor parents’ experiences. The popularity and success of survivordescendant dialogue workshops which have been conducted in numerous countries attests
both to the children of survivors’ need for clarity and the epiphanies that arise from open
communication about Holocaust trauma. Participation in such workshops can help to
remove ambiguities experienced by children of survivors created by their survivor parents’
non-verbal messages (Halasz, Nahum, Wein, & Valent, 2006).
The experiences of survivors and their descendants in relation to disclosure about
Holocaust trauma are very useful lessons for clinicians working with survivors of more
recent traumas. Not only can open discussion about trauma be a helpful and cathartic
experience for the victim/survivor (Finkelstein & Levy, 2006), but it can also help
minimise the transmission of traumatic impact through the generations by leaving the
survivors in better psychological shape to face the parental role. In addition, survivors can
be taught how their silence, which they may interpret as being protective of their children,
can actually be more harmful than being open and honest with their children. Telling their
children about their traumatic experiences, in the amount of detail that the child is,
developmentally, able to understand and absorb is in many cases the best course of action.
16.8. – Contemporary Needs
As was discussed in Chapter One, while the Nazi Holocaust ended over 60 years ago, it
continues to be a relevant area to research because of the emerging needs of the aging
survivor population. Many survivors are now so frail that they require specialised care
which can only be provided institutionally.
Holocaust survivors quite often find the
regimented atmosphere of a nursing home frightening, as it is reminiscent of life in ghettos
or camps. Staff in nursing homes with survivor residents need to be aware of these
© Janine Lurie-Beck 2007
348
“trigger” circumstances to help minimise negative symptoms among aging survivors.
Diminishing health can be particularly anxiety-provoking for survivors, given that one’s
survival during the Holocaust was heavily reliant on physical fitness (Williams, 1993), both
in terms of getting through the selection process and long-term ability to cope with the
physical hardships of hunger, disease and excessive labour.
With age [my survivor father] has become more anxious, especially over
insignificant matters and the memories of the camp are more with him
now as he gets older.
A child of survivors
If the theorised u-shaped curvilinear relationship between time lapse since the
Holocaust and symptoms experienced by survivors is true, then we are coming into (and are
to a large extent already in) a time when aging survivors are increasingly suffering from the
recurrence, and for some the first emergence, of depressive, anxiety, paranoid and intrusion
symptoms. Support for the notion of such a curvilinear trend was found for depression
levels among survivors via a meta-regression looking at time lapse since the end of the war
(see Chapter Eight, Section 8.7) although it was less straight forward for other variables.
When this upturn in symptoms is coupled with the anxiety provoked by failing physical
health and the negative connotations associated with that for survivors (Williams, 1993),
the aging process is potentially very traumatic for Holocaust survivors. Sixty years on, the
legacy of the Holocaust still requires the attention of professionals in the counselling and
aged-car spheres. The unique problems faced by aging Holocaust survivors as they move
into aged care facilities, as well as the increase or emergence of psychological symptoms,
has been acknowledged in recent print (e.g., an article in the Weekend Australian Magazine
on 31 March 2007 (F. Harari, 2007) and television media coverage (e.g., an episode of the
ABC [Australian Broadcasting Corporation] program Compass which addressed Holocaust
survivors with Alzheimer’s which aired on 22 October 2006 (see webpage –
http://www.abc.net.au/compass/s1748774.htm ). Indeed, many aged care facilities have
been forced to implement programs focussing on the specific needs of Holocaust survivor
residents (e.g., the Sir Moses Montefiore Jewish Home in Hunters Hills, Sydney [as
featured in the aforementioned ABC Compass program]).
Children of survivors also face difficult times as they are confronted by the decline
in their parent’s health and imminent death. The death of the majority of extended family
members during the Holocaust means that not only did survivors not have their parents to
© Janine Lurie-Beck 2007
349
turn to, but their children did not have their grandparents.
They did not have the
opportunity to watch their grandparents age or have the experience of mourning their death
of old age. This makes the death of survivor parents all the more difficult for their children
to deal with.
My mother’s death was extremely traumatic and I believe this was
because I never had any experience of death. I never knew any one in my
family who was older than my parents.
A child of survivors
Quite apart from the continuing needs of survivors and descendants of survivors of
the Holocaust, research into the effects of the Holocaust serves to highlight the current and
future needs of survivors of more recent genocidal actions such as those in the former
Yugoslav states, the African countries of Rwanda and Sudan and Asian countries such as
Cambodia and indeed any civilian population in a war-torn country such as Iraq or
Afghanistan. Given the long ranging effects that have resulted from the Nazi Holocaust,
clearly mental health professionals will be dealing with the survivors and descendants of
these more recent state-based traumas for many years to come.
16.9 – Clinical Significance and Applications
This thesis provides numerous lessons to be noted by clinicians providing therapy and other
support services to Holocaust survivors and descendants. Many have been learnt already
from clinical experience with survivor and descendant clients. However, it is important to
have this knowledge backed up by solid empirical evidence.
Clearly among the survivor generation itself, the Holocaust has been a central and
defining experience in their life, which continues to reverberate powerfully. Despite the
small sample of survivors who participated in the empirical study conducted for the current
thesis (and the associated statistical power issues) variables relating to elements of their
Holocaust experiences were found to explain up to 60% of the variation in their scores on
psychopathological and psycho-social adjustment variables. Time spent alone (without
family members) was clearly an important element of Holocaust trauma explaining up to
32% of variation in scores. Exploration of issues such as perceived abandonment, isolation
and separation anxiety are the obvious areas that spring to mind when contemplating
application to a clinical/therapeutic setting.
By breaking the Holocaust experience into traumatic elements, it has been possible
to identify the most traumatising aspects or combination of aspects of Holocaust trauma.
© Janine Lurie-Beck 2007
350
While the variables studied in the current thesis are specific to the Nazi Holocaust, many
elements are readily transferable to recent attempted genocides.
The identification of
highly traumatising sub-elements of Holocaust trauma can be used to help target limited
therapeutic/counselling resources for these more recent survivors. For example, the results
of the current research would suggest, that survivors who have no surviving family
members or who spent part of their traumatic experiences alone will, potentially, in general,
be more needing of therapeutic help because they are more likely to suffer from higher
symptom levels. Such survivors could easily be identified via initial screening interviews.
An educational/therapeutic module could be developed for survivors who become
parents to increase their awareness of the effects of their trauma on their parental modes.
Survivors could be taught about concepts such as developmental abilities to understand and
process details of a traumatic narrative.
They could be made aware of the potential
negative effects of ambiguous reactional behaviours (for example fear reactions for which
the stimuli is unclear to the child leading to potential generalised anxiety).
However, in cases where it was not possible to try to prevent the trauma-infused
parenting styles of survivors it is potentially possible to try to ameliorate its after-effects on
the children. Certainly both prior anecdotal literature as well as the findings of the current
thesis’s empirical study point to the influential role of ambiguous and emotive
communication about Holocaust experiences by survivor parents to their children.
Exploration of these communication modes and attempts to clarify ambiguous messages
can go a long way in helping children of survivors understand their family environment as
well as potentially alleviate the aftereffects of this family environment (such as increased
anxiety and problems within interpersonal relationships). The success of workshops in this
area alluded to previously suggest that group therapy sessions with children of survivors
have been and will continue to be useful therapeutic techniques.
In terms of influential psychological process mechanisms identified and studied
within the current thesis, therapists can be more aware of the affected cognitions such as
weakened world assumptions which may potentially be targeted during therapy with
survivors and descendants. Direct experience of the Holocaust and even the knowledge
that close family members experienced the Holocaust are readily understood reasons for a
weakening of world assumptions. Strengthening of world assumptions is an area that
clinicians may explore as an indirect way of reducing psychopathological symptoms.
© Janine Lurie-Beck 2007
351
Exploration of Holocaust issues should be one of the major tasks in any therapy
with descendant of survivor clients, with the caveat of course that the client is willing to
further explore feelings related to this major event. The analyses presented in the current
thesis point to the continuing importance of the Holocaust legacy in shaping descendant’s
identity and psychological health and functioning.
16.10. – Methodological Issues of the Current Thesis
The current thesis is comprised of two sections of research. The first involved metaanalyses of data from the literature and the second involved an empirical study. The main
point of caution that can be made about the meta-analytic results is the impossibility of
being able to say for certain that all relevant studies were included. While relatively
exhaustive efforts were made to locate both published and unpublished data, it is
acknowledged that some may have not been obtained. It should be noted that in addition to
the stringent efforts made to locate studies, calculations of Fail Safe Ns (which provide an
estimate of the number of studies with contradictory findings necessary to alter the metaanalytic result) were also made. The Fail Safe Ns gave the reader an idea of the stability of
the findings of the meta-analysis and the likelihood of the results being inaccurate if all
relevant studies were not included. Certainly it is acknowledged that even with metaanalyses, the state of evidence in the literature pertaining to a number of variables remained
unclear.
In many cases this is due to inconsistent operationalisation of variables
(especially demographic variables) and/or inadequate data.
There are a number of shortcomings of the empirical study conducted for the
current thesis which need reiterating. The most obvious point to make is the small sample
sizes on which analyses were based. When analyses depend on drilling down to sub-groups
of samples, it is impossible to maintain adequate power to conduct multivariate statistical
tests and so univariate and bivariate analyses were used. Obviously a multivariate approach
would be preferable to clarify the truly unique contribution of all the variables included in
the Model of the Differential Impact of the Holocaust across Three Generations. While
inter-relationships between demographic variables were carefully screened and controlled
for, where necessary, via relevant analysis techniques such as partial correlations and
ANCOVAs, it is acknowledged that this approach is not a replacement for a more robust
multivariate approach with a larger sample size. However, with the aging of the survivor
© Janine Lurie-Beck 2007
352
population the chances of obtaining large survivor samples for survey research, such as was
conducted for the current thesis, are rapidly diminishing.
Another issue pertaining to the sample of the empirical study is the degree to
which it is representative of the survivor and descendant populations. While attempts were
made to obtain participants from a range of different sources, it still must be acknowledged
that participants had to volunteer to participate and had to be either a member of a group or
exposed to media coverage regarding the study. The costs involved in attempting to obtain
a large enough random sample of survivors and descendants (for example via a much larger
community based study with screening questions for identification) were prohibitive for a
PhD research project. While such a method would be preferable, the findings based on
volunteer samples can also aid our understanding of the impacts of the Holocaust, as long
as the sampling methods used are taken into account by the reader.
The amount of information supplied by grandchildren of survivors about their
survivor ancestors made it impossible to analyse the effects of specific elements of their
survivor grandparents’ Holocaust experiences on their generation. Given the strength of
association between the elements of survivors’ experiences and the psychological health of
their children, it would be most desirable to be able to test if the ranking of these Holocaust
experience dimensions is repeated for the grandchildren of survivor generation. The fact
that the number of survivor grandparents and the number of child of survivor parents that a
grandchild of survivor has were both in the top three highest ranked demographic variable
in explaining that generation’s symptom levels points to the continuing strong influence of
survivor ancestry one generation further removed.
The failure to obtain non-Jewish survivors (or descendants of non-Jewish survivors)
was also particularly frustrating, given the efforts to obtain such participants by the author
(refer to Appendix H for the organisations contacted). If a group of non-Jewish survivors
had been obtained, in particular those who were interned in camps for political opposition
to the Nazi regime, or a group of survivors who had survived in hiding by seeking out there
own food and shelter, an analysis of the importance of active versus passive roles could
have been conducted. Given the pattern of study findings with other populations pointing
to the importance of active versus passive roles in traumatic experiences it is believed this
could have proved an important distinction among Holocaust survivors. It is still believed
by the current author that non-Jewish survivors who were persecuted because of active
© Janine Lurie-Beck 2007
353
resistance to the Nazi regime will evidence lower symptom levels than those persecuted
purely based on their ethnicity or religions (as was the case for Jews and Gypsies). It
would also be of interest to compare survivors who were in hiding based on whether they
were hidden by others and dependent on them for food and safety, or whether they survived
in hiding and sought food and necessities for themselves.
In an attempt to reduce the size of the questionnaire booklet to be completed by
participants of the empirical study, only some subscales of a number of measures were
used. Specifically, this refers to the use of only the cohesion and expressiveness subscales
of the Family Environment Scale, the Fostering of Autonomy Scale of the Parental
Attachment Questionnaire, and the belief in world benevolence and belief in world
meaningfulness subscales of the World Assumptions Scale. Given the small sample sizes
obtained for the empirical study, and the impossibility of multivariate analyses with the
obtained data, the addition of extra variables/subscales would have made the ensuing
collection of univariate analyses very jumbled indeed. The subscales were chosen to
narrow in on specific facets of survivor family experiences as outlined in the literature. It
must be said that the main focus was not to make an exhaustive list of the influential
psychological processes and transmission processes that could impinge on survivor and
descendant psychological health. Rather, the aim was more to develop a “bare bones”
model of these processes, with the focal point being more on the role of demographic
variables in moderating the effects and transmission of Holocaust trauma. Therefore, it is
acknowledged that there are possibly additional variables that could be considered
influential psychological processes or family interaction/transmission processes that are
worthy of consideration, in obtaining a more complete picture of the processes which lead
to intergenerational trauma transmission.
While discussing the measures used in the empirical study, a cautionary note must
be reiterated in relation to three of the scales used. The internal consistency reliabilities for
the empirical study sample for the Posttraumatic Vulnerability Scale (PTV), the cohesion
subscale of the Family Environment Scale (FES) and a few of the subscales of the
Holocaust Communication Questionnaire (HCQ) were lower than would have been desired.
It must be acknowledged that previous research indicated both the PTV and the cohesion
subscale of the FES to be of sound reliability, while the HCQ had not been analyses for
reliability prior to its use in this study. Therefore, the knowledge of their lower reliability
© Janine Lurie-Beck 2007
354
in the current study could not have been known apriori. Because of the exploratory nature
of the empirical study, and the central role hypothesised by communication about the
Holocaust on the part of survivors to their children, it was felt that to leave the HCQ could
not be left out of the analysis. The central role that family cohesion was hypothesised to
play (and was discovered to play) in terms of trauma transmission meant that the omission
of the FES cohesion subscale would have left its role uncovered. The PTV measure upon
examination of its items was found to have high face/construct validity in relation to the
specific fears regarding safety so often reported in the Holocaust survivor literature.
However, it is obviously acknowledged that caution should be used when considering using
the PTV in future research. Perhaps if another measure could be located or created that
examined similar issues, the two measures could be used in unison to provide a back up
source of data if the PTV was found to have low reliability in a given study. One of the
main problems of the HCQ is the small number of items which load onto the subscales
(ranging between 1 and 5 items). There is certainly scope to expand the measure, to
perhaps pick up on more of the subtle nuances of communication that are not currently
measured by the scale. To increase the number of items loading on each subscale can only
have a positive influence on its internal consistency. Certainly replication of the results
obtained in this study using an alternative measure of family cohesion would not be hard
given the number of measures of this construct in existence.
It was originally envisaged that the empirical study would be able to examine the
relationships between all the variables in the Model of the Differential Impact of the
Holocaust across Three Generations by having a large enough sample of survivor families
for which at least one member from each generation had participated. Unfortunately,
despite the efforts of the author and many participants to encourage family members to
participate in the study, there were an insufficient number of families for which three or
even two generations were represented in the sample. It was still possible to examine the
effects of ancestral demographic variables, as these details were provided by participants
themselves. However, it was not possible to examine the relationships between ancestor
and descendant scores on psychological impact and influential psychological process
variables from the Model of the Differential Impact of the Holocaust across Three
Generations. While relationships between parental pathology and offspring pathology had
been established within the general population and the Holocaust survivor population in
© Janine Lurie-Beck 2007
355
previous research (Keinan et al., 1988; Major, 1990; Schwartz et al., 1994; Yehuda et al.,
2001), it would have been desirable to replicate this finding within the current thesis’
empirical study data set.
However, despite the shortcomings noted in this section, this thesis adds
significantly to the literature with regard to the impact of Holocaust survivors and
descendants. These methodological issues do little to detract from the validity of the
findings reported in the current thesis. They largely reflect additional variables that require
consideration rather than suggesting the findings based on the variables assessable are not
relatively robust.
16.11. – Future Research Directions
While the research in the current thesis has filled some of the gaps in the available
knowledge about the impact of the Holocaust on survivors and descendants, there is still
much that could be uncovered. It seems that the more research that is conducted in this
area, the more questions arise.
An area that could not be addressed in the current thesis but which is considered
potentially important by the current author is an examination of active versus passive roles
during the Holocaust on post-war psychological health. Delineation of a range of active
versus passive roles and experiences during the Holocaust is possible and the Holocaust
provides an ideal opportunity to examine this issue further.
The area of communication about the Holocaust can certainly be researched in more
depth.
Now that the grandchildren of survivors are coming into adulthood in large
numbers, the issue of how they learn about their survivor grandparents’ experiences and the
impact of this knowledge needs to be addressed. This generation will be the last to have
direct contact with the survivor generation; indeed some of the grandchildren generation
may not have had direct contact with their survivor grandparents. Is the turning point in the
measurable transmission of Holocaust trauma the point at which direct contact with the
survivor generation ends? Are there differences between grandchildren who had contact
with their survivor grandparents and who perhaps heard the Holocaust stories first hand,
and grandchildren who merely found out about a grandparents’ Holocaust experiences after
their death? Does the knowledge of ancestral impact have as strong an affect, if the
survivors’ grandchildren never had any direct contact or interaction with their survivor
grandparents? Krasnostein (2006) said of grandchildren of survivors:
© Janine Lurie-Beck 2007
356
We are the pivotal generation, willing to look at the effects of the
Holocaust with a different and perhaps more distant view than our
parents, the second generation, and our survivor grandparents are able to.
However, one level of tension exists because even though there is more
distance, we are still so close because it happened to our family.
If discernible differences between children and grandchildren of survivors and the
general population are present, at which point do these differences disappear and the
descendants essentially “normalise”? Only when this question is answered can we truly
know the full extent of the reverberating effect of the Holocaust.
The construct of world assumptions as explained in Janoff-Bulman’s (1992) Theory
of Shattered Assumptions is one that is currently under-researched with the survivor and
descendant populations. In particular, it would be of interest to explore the relationship
between survivor parent and offspring’s’ world assumptions. To what extent is a person’s
world view inherited from their parents, and when that view is largely negative, is it
possible for the children to revise their own world view and to what degree? Is the
relationship between parental and offspring scores on pathology measures mediated by
strength of belief in world assumptions?
Given that survivors’ loss of family was revealed as very important in both their
own and their children’s psychological health, a future study could perhaps address this
issue more directly by using a measure of unresolved mourning. Such a measure was
considered for the empirical study but abandoned due to fears that response bias would
render the results meaningless. If a measure of response bias could be incorporated, then it
would be useful to explore this issue in more depth with survivors.
The Model of Differential Impact of Holocaust Trauma across Three Generations
developed and tested over the course of the current thesis could be readily adapted and
applied to survivors of more recent large scale state-based traumas.
Many of the
demographic variables (such as loss of family, age, country or perhaps region of origin) are
readily transferable, while others (such as the nature of experiences) would have to be
moulded to fit the specific trauma set being examined. The focus should be on determining
the specific elements of the trauma itself and how these may differ. The main aim of this
process is to identify the most vulnerable demographic sub-groups so that psychological
help, if limited, can be more effectively targeted at those likely to need it the most. The
model can be used to help inform both survivors and mental health professionals the way in
© Janine Lurie-Beck 2007
357
which parent-child and whole family interaction patterns can be affected with the hope that
this awareness can help minimise the transmission of trauma to the children of those
affected.
16.12. – Conclusions
To the first question: “How could it have happened?” we respond with
guilty silence and to the last one: “Could it happen again?” we nod in
shame.
Kellerman (2006b, p. 30)
From examining the research that has been conducted to date with Holocaust survivors and
their descendants in detail, it is apparent that many opportunities to gain knowledge were
squandered. Minimal research was conducted in the immediate aftermath of the Holocaust
and it was not until decades later that methodologically sound studies were conducted.
Data on factors that have since been shown to be influential in determining post-Holocaust
adjustment were not consistently collected or analysed. Because of this, research such as
that conducted for the current thesis, has had to play “data collection catch-up” with a
rapidly diminishing population before the opportunity to study survivors is lost forever.
We have not even “scratched the surface” when it comes to understanding the role of
factors such as cultural differences, country of origin, reasons for persecution and the
hypothesised protection of active roles have on post-Holocaust adjustment.
We live in a world where men, women and children continue to be subjected to state
or community-based traumas. It is imperative that we take the lessons learned from the
Holocaust survivor population and use them as guiding principles in helping the survivors
of more recent traumas. In the many areas where our knowledge regarding the Holocaust is
lacking, we need to conduct methodologically sound, large scale, longitudinal studies with
survivors of more recent-state based traumas, so that we can more accurately understand
their long-term prognosis and the factors that underpin it.
We need to ensure that immediate and continuing psychological help and
monitoring is available to these survivors.
We need to approach this task with the
knowledge of how the impacts of these traumas can be passed on through the generations.
If we have been unable to help the survivors to the level we would like, then we need to be
willing to educate the subsequent generations to help them understand their parents’ and
ancestors’ behaviours and attitudes so the likelihood of dysfunctional parenting patterns
being repeated for many generations can be minimised.
© Janine Lurie-Beck 2007
358
The world needs to understand that the effects of state-base trauma are long-lasting
both for the individuals directly effected as well as their descendants, for potentially more
generations to come. They also need to acknowledge that not all people will react and/or
recover from such experiences in the same way. It is the role of researchers in the areas of
psychology, psychiatry and social work to educate the world as to the nature of this
reverberating traumatic impact in the hope that more attention will be focussed on large
scale persecution and traumatisation of civilian populations and that more will be done to
stop and prevent it. Unfortunately, the effects of recent large scale community persecutions
and “ethnic cleansings” (such as in the former Yugoslav states, Rwanda, Cambodia and
very recently in Darfur in the Sudan) are already providing psychologists with an evergrowing group of survivors and potentially traumatised descendants.
While the Nazi Holocaust has provided us with evidence of long-lasting negative
effects of survival, there is also something heartening to take from encounters with
survivors. That is that while they have suffered psychologically because of their Holocaust
experiences (to differing degrees) they have also to a large extent been able to lead
relatively successful and productive lives in their post-Holocaust years.
This fact
demonstrates the strength of the human spirit to overcome even the most horrific traumas.
After a while you find you can smile again and you live again and you
think to yourself I am still alive…
Jozefa Lurie, a survivor
Figure 16.2. Polish prisoners in Dachau toast their liberation from the camp (circa April/May 1945)
Source: United States Holocaust Memorial Museum Photo Archive (Photograph # 83818)
Note. The views or opinions expressed in this thesis, and the context in which the image is used, do not necessarily
reflect the views or policy of, nor imply approval or endorsement by, the United States Holocaust Memorial Museum.
&c