Analysis and write-up of qualitative data

Analysis and write-up of
qualitative data
• Transcripts
• Observations of nonverbals
• Contextual information
– Specifics of contact
– Historical info
• Verbatim of written material, speeches, etc.
• Info on observer
The problem with qualitative
• Often a mass of data with no rules for the
researcher to apply
• There are no pre-defined categories,
dimensions, etc. to tally up
• The language used by interviewees, etc. has
multiple meanings and interpretations
– Analysis is inherently subjective
Types of analysis
• Dialectical analysis
– Search for the most powerful conflicts in the
• Metaphor analysis
– Identifiy metaphors and how they vary
• Fantasy theme analysis
– Looks at the stories shared among people
Grounded theory and analysis
• Grounded theory is the most common form
of organized qualitative data analysis
– Developed as a means to bring rigor and
therefore greater acceptance of qualitative
methods in social science (sociology)
Grounded theory is:
• “An inductive, comparative, and interactive
approach to inquiry that offers several openended strategies for conducting emergent
– Charnaz
How did grounded theory
• 1967 Glaser and Strauss book The
Discovery of Grounded Theory
• 1990, 1998 Strauss and Corbin prescriptive
form with predetermined categories and
concerns about reliability and validity
• 2000 Charmaz introduces “Constructivist”
Educational Research 2e: Creswell
Key characteristics of grounded
theory designs
A process approach
Theoretical sampling
Constant comparative method
Series of coding and recoding data
Theory generation
Educational Research 2e: Creswell
Role of the researcher
• The researcher serves as both data collection
instrument and analysis method
– The two are linked—adjust data collection as analysis
leads in certain ways
• The researcher is trying to develop theory—
– Theory built upward from the data (“grounded”), not
downward from premises to theory to hypotheses to
– Keeps theory close to the data
– Avoids imposing inappropriate theory on ‘reality’
• Three stages:
– Open coding
• Coding line-by-line original data into codes that the
researcher determines to be valuable
– Extremely subjective
– Axial coding
• Combining original codes into major categories and
defining subcategories and their relations to the
– Restricted coding
• Identifying relationships among codes and
Zig-zag approach to data
collection and analysis
Data Collection
Data Analysis
Close to Saturated
More Refined
Educational Research 2e: Creswell
Saturation of
Coding for what is happening
So part of it has been good because
I can see that I'm not the only one
that has good days and bad days,
everybody does.
They might not be physical, as
much as psychological, but
everybody has kind of good days
and bad days, as moods and things
Example from Charnaz
Identifying a
Recognizing other
people’s good
and bad days
Qualifying their good
and bad days
Viewing good and
bad days as
Comparing Statements: Sara Shaw—
Taking a broader view beyond self
So part of it has been good because I can
see that I'm not the only one that has good
days and bad days, everybody does.
They might not be physical, as much as
psychological, but everybody has kind of
good days and bad days, as moods and
things too
Seeing beyond self
Discerning the content of
good and bad days
Comparing statements: Nancy Swensen dealing with
her illness on a bad day and her mother with
Alzheimer’s—Being caught in chaos
And if I’m trying to get dinner
ready and I’m already feeling bad,
she’s in front of the refrigerator.
Then she goes to put her hand on
the stove and I got the fire on.
And then she’s in front of the
microwave and then she’s in front
of the silverware drawer. And-and
if I send her out she gets mad at
me. That’s when I have really a
really bad time.
Making a bad day
Escalating chaos
See also,
Arthur Frank (1995):
“The Chaos Narrative
Comparing responses to bad days:
Dealing with bad days
We’re [a friend who has
multiple sclerosis] kind of like
mutual supporters for each
other. And when she has her
bad days or when we
particularly feel “poor me,” you
know, “Get off your butt!” You
know, we can be really pushy to
each other and understand it.
Having bad days
Disallowing self-pity
Issuing reciprocal
Taking the criticism
Realizing that once bad days have become
good days—John
What used to be bad days
[laughing] now are good days
…but the quality of things, I think,
is declining, you know, from , say
a couple of years ago when I didn’t
think about it that much. And
there would be isolated days when
I had a lot of congestion and things
like that. But that’s all.
Shifting criteria of
good and bad days
Defining declining
Comparing past and
• “Eventually, new data . . . add little to the
development of new descriptive categories.
At this point, the categories are considered
“saturated.” The researcher then reviews the
theoretical memoranda and conceptualizes
higher level (more abstract) generalizations
that subsume the initial set of categories yet
are grounded in them.”
– P. 282
• “Memos are notes the researcher writes
throughout the research process to elaborate
on ideas about the data and the coded
categories. In memos, the researcher
explores hunches, ideas, and thoughts, and
then takes them apart, always searching for
the broader explanations at work in the
Educational Research 2e: Creswell
• The grounded theory generated in this
manner reflects the researcher’s
development of categories of meaning and
the relationships she perceives among them.
It should also provide some explanation for
those relationships.
– Usually not “higher-order” theory, but “midlevel” theory
• One test of the theory is to have the
interviewees react to it. Does it make sense
to them? Do they think it reflects the world
as they see it?
• The theory developed through this method
should be compared to the scholarly
literature to see how it fits within the field.
Does it add, support, contradict? The
researcher should provide his thoughts
about how the grounded theory should be
interpreted with regard to existing theory.
• Write-up of a grounded theory study
follows more of a narrative format than
traditional quantitative research reports.
While the concerns of the researcher that
led to the study usually open the piece, they
are often more closely tied to the researcher,
personally, than in quantitative studies.
• The discussion of the scene to be studied is
often quite detailed, while a theoretical
literature review is limited or absence. The
methods and results are more integrated and
the author will often discuss the evolution
of her thinking over time, tying it to
particular findings in narrative format—
over time.
• Evidence is often provided in direct quotes
and examples, and the development of the
researcher’s thinking over time is
commonly revealed.
• The latter portion may discuss a comparison
of the grounded theory with existing theory
to a greater extent than is often the case.
Write-up and presentation
• Provide a detailed description of the data
collection methods
– Describe the subjects and context carefully
Evaluating a grounded theory study
• Does the researcher gather extensive data so
as to develop a detailed conceptual theory
as well saturated in the data?
• Does the model emerge through phases of
coding? (e.g. initial codes to more
theoretically oriented codes or open coding
to axial coding to selective coding)?
Educational Research 2e: Creswell
• Does the study show how the researcher
validated the evolving theory by comparing
it to the data, examining how the theory
supports or refutes existing theories in the
literature, or checking theory with
– Creswell
Evaluating a grounded theory study
• Is there an obvious connection between the
categories and the raw data?
• Is the theory useful as a conceptual
explanation for the process being studied?
• Does the theory provide a relevant
explanation of actual problems and a basic
• Can the theory be modified as conditions
change or further data are gathered?
Educational Research 2e: Creswell