1 British Journal of Social Psychology (2011) ! C 2011 The British Psychological Society The British Psychological Society www.wileyonlinelibrary.com Brief report What is the best model for girls and boys faced with a standardized mathematics evaluation situation: A hardworking role model or a gifted role model? C´eline Bag`es∗ and Delphine Martinot∗ Clermont Universit´e, Universit´e Blaise Pascal, Laboratoire de Psychologie Sociale et Cognitive, UMR CNRS, Clermont-Ferrand, France Same-gender role models are likely to improve girls’ math performance. This field experiment examined whether the explanation given for a role model’s success also influence children’s math performance. Fifth graders were presented with a female or a male role model before a difficult math test and were informed about the cause of his/her math success (effort vs. ability vs. no explanation). The results showed that the gender of a hardworking role model did not influence math performance. In contrast, when the role model’s success was not explained or explained by abilities, children performed better with the female role model than with the male role model. The hardworking role model and the female role model allowed reducing stereotype threat among girls. Chelsea is in fifth grade and does her math homework. Her neighbours, William and Emma, both brilliant sixth graders in mathematics, come to visit her and talk about their success in this domain. William talks about the important effort and hard work he regularly puts into it, whereas Emma mentions her gift. Which of these two children is likely to have the most influence on Chelsea’s progress in math? As a girl, Chelsea is likely to be more inspired by another girl than by a boy (Lockwood, 2006). Therefore, Emma is expected to be the best role model to influence Chelsea’s math performance. However, individuals do not necessarily consider the role model’s gender as the most relevant information to make expectations about their own future success (Bandura, 1997; Javidan, Bemmels, Devine, & Dastmalchian, 1995). Therefore, is Emma really the best role model for Chelsea, while Chelsea has information beyond gender, as the reason of her neighbours’ math success? This experiment aims to answer to such a question in studying whether the explanation given for a role model’s math success moderates the role model’s gender effect on children’s math performance. ∗ Correspondence should be addressed to C´eline Bag`es and Delphine Martinot, LAPSCO, 34 avenue Carnot, 63000 ClermontFerrand, France (e-mail: [email protected], [email protected]). DOI:10.1111/j.2044-8309.2010.02017.x 2 C´eline Bag`es and Delphine Martinot Examining who may be the best role model for improve Chelsea’s math performance is an especially relevant issue. Indeed, although girls actually perform as well as or better than boys in math in elementary school and in the first years of junior high school (see Else-Quest, Hyde, & Linn, 2010, for a review, and see PISA,1 TIMMS,2 for surveys), they continue to perform less well on standardized math tests when the evaluative context makes salient the stereotype regarding girls’ poor math abilities (e.g., Huguet & R´egner, 2007; Keller & Dauenheimer, 2003), particularly in France (see PISA). In such a threatening situation, could young girls benefit from exposure to role models renowned for their success in math? And overall, who is the best role model for children in math? As comparison targets (Gibson, 2004), successful role models may be considered as a source of information to make expectations about one’s own future success (e.g., Lockwood & Kunda, 1997), and may exert a positive impact on motivation (Lockwood, Jordan, & Kunda, 2002), self-evaluation (Lockwood & Kunda, 1997), and performance (Earley & Kanfer, 1985), but only when they are perceived as attainable role models. Indeed, Lockwood and Kunda (1997) showed that students were more inspired by an advanced student than by a same-year student, as they realized that it was too late for them to accomplish in school what this outstanding student had managed to achieve. Moreover, slightly upward comparison targets are likely to be inspirational and lead to academic progress when individuals feel similar to them (see Dijkstra, Kuyper, Buunk, Van der Werf, & Van der Zee, 2008, for a review). Thus, fifth graders are likely to perceive a sixth grader (i.e., slightly upward comparison) as a relevant role model because he/she is similar to them in terms of academic interests, but not self-threatening because his/her achievements seem attainable. Moreover, interesting benefits of role models have been shown in the field of stereotype threat. Marx and Roman (2002) observed among adults that a successful female role model in math (i.e., a counter-stereotypical model) minimizes the impact of gender stereotypes by permitting the women to think that they, too, can be successful in math. In contrast, presenting women with a male role model (i.e., a stereotypical role model) reminds them how difficult it is for them to be as successful as men in math: such a model could demoralize them and lead to poorer performance. Because boys benefit from a positive gender stereotype concerning math ability, the role model’s gender should be less relevant for them and not influence their math performance. Therefore, we suggest that girls should perform as well as boys in a difficult math test after exposure to a successful female role model in math but less well than boys if the role model is male (H1). However, we suggest that the explanation given for the role model’s math success (effort or abilities) may create a motivational framework, which is likely to moderate the impact of the role model’s gender on children’s math performance. Indeed, this explanation concerning the role model’s math success may affect children’s beliefs regarding the link between effort and performance. Success attributed to a person’s unrelenting effort and regular work is a controllable explanation, whereas success attributed to abilities is not (Weiner, 1985). Perceiving the negative content of a stereotype as being controllable minimizes the effects of stereotype threat on women’s performance. Thoman, White, Yamawaki, and Koishi (2008) showed that women 1 PISA: Organization for Economic Co-operation and Development, Programme for International Student Assessment (http://www.pisa.oecd.org) 2 TIMMS: Third International Mathematics and Science Study (http://timss.bc.edu/timss1995.html) Role model and student’s math performance 3 performed better on a math test when the superiority of men in math was explained by their more intensive efforts (a controllable factor) rather than by a biological difference (an uncontrollable factor). Good, Aronson, and Inzlicht (2003) also reported that female college students had significantly higher standardized math test scores when they were mentored by college students to consider intelligence as malleable or to attribute academic difficulties in the seventh grade to the novelty of the educational setting. Thus, students who receive information to consider intelligence as a result of controllable factors such as effort and hard work perform better and are more motivated than students who think of intelligence as a gift or a fixed trait which is unlikely to develop with learning (Dweck, 1999; Mueller & Dweck, 1998). Comparison information also leads to perceive one’s performance as controllable or not. Indeed, upward comparison information that indicates how to improve oneself (i.e., to exert high effort) reinforces an individual’s perception of control over his/her own performance (Van Yperen, Brenninkmeijer, & Buunk, 2006). The opposite is observed when upward comparison information leads individuals to believe that effort will not improve their performance (Van Yperen et al., 2006). Accordingly, children exposed to a description of a successful role model renowned for his/her gift in math may reason that math ability is innate, and may infer that working harder would be insufficient to succeed because their math performance is not controllable. On the contrary, children who learn that the math success of a role model is the result of a long struggle and hard work may reason that it is possible for them to overcome difficulties and achieve success, and consequently may be more willing to engage in processes (e.g., working harder) to achieve successful outcomes. Therefore, in a second hypothesis (H2), we suggest that boys as well as girls, exposed to a role model whose math success is explained by his/her gift, should perform worse in math than children exposed to a role model renowned for his/her effort, whatever the role model’s gender. In contrast, when the role model’s success in math is not explained, the role model’s gender may become relevant: only a female role model will enable girls to succeed just as well as boys in math, without impairing the performance of the latter. Method Participants In all, 405 French fifth-grade children (196 girls and 209 boys), mean age 10 years 7 months (SD = 5 months), participated in the study with their parents’ consent. The children attended urban and rural state schools (22 classes) selected in order to reflect a wide variety of social backgrounds. The experiment was presented to the parents as a ‘study of children’s performance at school’ and they were informed that the children’s data would remain confidential. We also obtained each pupil’s math grades before conducting our study. The following experimental design was used: 2 (Pupils’ gender) × 2 (Role model’s gender) × 3 (Explanation for the role model’s math success: effort, ability, or no explanation), and the children were randomly assigned to the role model’s gender by success explanation conditions. Procedure A female experimenter randomly divided each class into mixed-gender groups of 8 to 14 students because math evaluations are performed in a mixed gender context in France. To avoid the ‘teacher effect’ (Nye, Konstantopoulos, & Hedges, 2004), the pupils’ teachers were not present during testing. The experimenter distributed booklets 4 C´eline Bag`es and Delphine Martinot containing the experimental manipulation and the math test. The pupils in all the conditions were asked to read a short text about a sixth grader’s success in math (Marc or Marie depending on the role model’s gender) before completing a standardized math test. For all the pupils, Marc/Marie was described as a successful student in math. In the hardworking role model condition, this math success was explained by the sixth grader’s regular efforts and hard work. In the gifted role model condition, this success was explained by his/her talent for math. In the unexplained condition (i.e., only the skilled role model), no reason was given for his/her math success. The experimenter then administered a math test that included 10 exercises from the French national standardized math evaluation, usually administered to sixth graders at the beginning of the school year. This test was feasible but difficult for fifth graders. All of the exercises were selected because they represented items on which the performance of French sixth-grade boys is 8% to 17% better than that of girls (French Department of Education, 2007). The instructions given to the pupils were identical in all respects to those used when the test is formally administered. The pupils had 1 hour to complete the test. A grader, blind to gender and condition, awarded 1 point for a correct answer and 0 for an incorrect answer or no answer. The highest possible score was 10. After completing the test, to check that the pupils share the role model’s interest in math, they had to rate the importance they accorded to math on seven items (e.g., ‘I study math because I know how useful it is’) from the Mathematics Attitudes Scale (Vezeau, Chouinard, Bouffard, & Couture, 1998). The reliability level computed on the seven items was acceptable (Cronbach’s ! = .83). Each item was scored on a 5-point scale ranging from 1 (not at all) to 5 (very much). Finally, we verified that the children were able to distinguish between ‘working hard’ and ‘being gifted’. The pupils had to match these concepts with a number of positive items (e.g., ‘work a lot’, ‘have great ability’, ‘spend a lot of time learning and doing exercises’, etc.). This manipulation check was administered at the end of the experiment, just before a thorough debriefing. Eight participants (2 girls and 6 boys) who failed to distinguish between the concepts of hard work and innate talent were eliminated, thus leaving 397 participants for the final analysis (194 girls and 203 boys). Results The preliminary analyses showed no effect of the pupils’ previous math grades on the different measures (math importance and math test performance), all Fs < 1 ns,3 and therefore it was not included in subsequent analysis. We then conducted a 2 (pupil’s gender) × 2 (role model’s gender) × 3 (explanation of the role model’s math success) ANOVA (Analysis of Variance) on each of the measures with all the variables treated as between-participant factors. Importance attached to math3 The analyses did not show any effect of our variables on math importance, all Fs < 1 ns. However, as expected, on average, the pupils cared about their performance in math (M = 3.88, SD = .81), as confirmed by the individual one-sample t-test compared to the value 3 (midpoint), t(330) = 19.74, p < .001. 3 Sixty-one participants did not complete this measure because their schools’ head teachers refused to give authorization. Role model and student’s math performance 5 Table 1. Mean math test performance as a function of the role model’s gender and the role model’s success explanation Female role model Male role model Role model’s success explanation n M SD n M SD Effort Abilities No explanation 75 66 68 5.79a 4.74a 6.11a .27 .28 .29 72 63 53 6.05a 3.99b 4.91b .27 .28 .31 Note. Within a line, means without a common subscript differ at a significance level of at least p " .05. Standardized math test performance The ANOVA showed a main effect of pupils’ gender, F(1, 385) = 24.44, p < .001, # = .24. The boys (M = 5.83, SD = .16) performed better than the girls (M = 4.70, SD = .16) in the math test. The main effect of the role model’s gender was significant, F(1, 385) = 6.11, p < .05, # = .13. The pupils performed better with the female role model (M = 5.55, SD = .16) than with the male role model (M = 4.98, SD = .16). The main effect of the explanation given for the role model’s success was also significant, F(1, 385) = 30.39, p < .001, # = .27. The pupils exposed to the gifted role model presented the lowest performance (M = 4.37, SD = .19) compared to the other two conditions (no explanation of the role model’s success, M = 5.51, SD = .21; hardworking role model, M = 5.92, SD = .19). Although the three-way interaction was not significant (F < 1, ns), all the two-way interactions were significant confirming partially H2. The Role model’s gender × Explanation for the role model’s math success interaction was significant, F(2, 385) = 3.75, p < .05, # = .14, demonstrating that, as expected, the explanation given for success moderated the impact of the role model’s gender on math performance. When the role model was presented as hardworking, the children performed just as well with a female (M = 5.79, SD = .27) and a male (M = 6.05, SD = .27) role model, F < 1, ns (Table 1). In contrast, when the role model’s success was not explained or explained by abilities, both boys and girls performed better with the female role model (gifted role model, M = 4.74, SD = .28, ‘no explanation’ role model, M = 6.11, SD = .29) than with the male role model (gifted role model, M = 3.99, SD = .28, ‘no explanation’ role model, M = 4.91, SD = .31), F(1, 385) = 4.74, p < .05, # = .10, and F(1, 385) = 8.40, p < .01, # = .14, respectively. The Pupils’ gender × Role model’s gender interaction confirmed the virtue of a female role model for both the boys and the girls, as expected in H1, F(1, 385) = 8.93, p < .01, # = .15 (Table 2). The girls scored (M = 5.32, SD = .22) just as well as the boys (M = 5.77, SD = .22) with the female role model, F(1, 385) = 2.04, p > .10, ns, but they Table 2. Mean math test performance as a function of pupils’ gender and the role model’s gender Girls Role model’s sex Female role model Male role model Boys n M SD n M SD 104 88 5.32a 4.08b .22 .24 105 100 5.77a 5.89a .22 .23 Note. Means without a common subscript differ at a significance level of at least p " .05. 6 C´eline Bag`es and Delphine Martinot Table 3. Mean math test performance as a function of pupils’ gender and the role model’s success explanation Girls Boys Role model’s success explanation n M SD n M SD Effort Abilities No explanation 70 63 59 5.29a 4.28b 4.54b .27 .28 .29 77 66 62 6.55a 4.45b 6.48a .26 .28 .29 Note. Within a column, means without a common subscript differ at a significance level of at least p " .05. underperformed (M = 4.08, SD = .24) compared to the boys (M = 5.89, SD = .23) with the male role model, F(1, 385) = 29.64, p < .001, # = .27. Finally, the Pupils’ gender × Explanation for the role model’s math success interaction was significant, F(2, 385) = 4.88, p < .01, # = .16. The best role model for the girls was the hardworking role model since they performed best after exposure to this role model (M = 5.29, SD = .27) compared to the gifted role model (M = 4.28, SD = .28) or the ‘no explanation’ role model (M = 4.54, SD = .29), F(1, 188) = 4.72, p < .05, # = .16. The boys’ math performance was lowest when they were exposed to the gifted role model (M = 4.45, SD = .28) compared to the boys exposed to the ‘no explanation’ role model (M = 6.48, SD = .29) or the hardworking role model (M = 6.55, SD = .26), F(1,203) = 34.03, p < .001, # = .38, (Table 3). Discussion The results of this research are encouraging, particularly because they were obtained in a field context rather than in a laboratory, with just a brief description of a successful pupil. Two of the role models were particularly beneficial: a hardworking role model and a female role model. According to Hypothesis 1, a successful female role model could improve girls’ performance in a math test by overcoming the negative effects resulting from gender stereotype, but without impairing boys’ math performance. These effects obtained with pupils replicated those reported by Marx and Roman (2002) with adults. Moreover, and as expected, the explanation given for success moderated the impact of the role model’s gender on math performance. Thus, when the reasons for success were not identified or seemed uncontrollable, the female role model was a better role model than the male role model. Indeed, when the role model’s success was not explained or explained by abilities, both boys and girls obtained a better math score with the female role model than with the male role model. In contrast, the role model’s gender might be neglected if his/her success was linked to controllable factors as regular efforts. When the role model was presented as hardworking, the children’s math performance did not depend on the role model’s gender. The present results also showed that a gifted role model led children to have their worst math performance. Then, a gifted child should be avoided as a role model for other children. When the role model was simply presented as successful in math (i.e., no explanation of the role model’s success), he/she seemed to inspire the boys as much as a hardworking role model, but he/she seemed to inspire the girls less than a hardworking role model. We suggested that the role model for which no explanation for his/her math success was given, offered the girls no information to help them to succeed in this threatening domain. This lack of information might lead girls to think Role model and student’s math performance 7 that they were inferior in math compared to this model and it always will be so. In contrast to the girls, the boys would not need guidance on how to succeed because they have a good reputation in math (Steele, 2003). In sum, only the hardworking role model seems to be beneficial for both gender groups. However, future research should attempt to understand how a role model may cause pupils to doubt themselves and their own success, thus leading them to ‘give up’, or, on the contrary, to increase their motivation and effort. We suggest that a gifted role model leads children to perceive less control over their performance or to diminish their self-efficacy (e.g., Bandura, 1997; Bouffard & Bordeleau, 1997), resulting in less efforts on the math task at hand. In contrast, a hardworking role model may increase perceived control and self-efficacy leading to greater efforts on the math task at hand. In spite of its limits, the main contribution of this study is to show that the explanation given for a role model’s success moderates the role model’s gender effect on children’s math performance. Previous research has shown that girls were more inspired by another girl than by a boy (Lockwood, 2006), and only a successful female role model in math could minimize the stereotype threat on girls’ math performance (e.g., Marx & Roman, 2002). The present results showed that a role model, regardless of his/her gender, who has succeeded in math thanks to regular hard work helps girls to obtain their best math score and thus to reduce the stereotype threat, even if the performance difference between the boys and the girls did not disappear with this hardworking model. Moreover, our study confirms that an explanation of personal success attributed to a controllable factor such as effort has a more positive effect on individuals’ performance than an explanation of personal success based on uncontrollable factors such as talent or biology (e.g., Dweck, 1999; Good et al., 2003; Thoman et al., 2008). Because the present results were obtained in a field context, they contribute to advise that everyone must be careful in explaining the success of others, especially when parents and teachers praise the success of persons who they wish the children would take as role models. When they are presented as hardworking, both male and female role models may be chosen to help girls to succeed in math. Thus, male classmates and brothers, renowned for their efforts, would be able to inspire girls. Acknowledgements We express our gratitude to C´eline Darnon and Sandrine Redersdorff for their helpful advice and comments on earlier drafts of the manuscript. References Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman. doi:10.1002/ 9780470479216.corpsy0836 Bouffard, T., & Bordeleau, L. (1997). Croyances de contrˆ ole et rendement scolaire chez des e´l`eves francophones du primaire au Qu´ebec. [Control beliefs and academic achievement of Frenchspeaking elementary school children in Quebec]. International Journal of Psychology, 32, 231–245. Dijkstra, P., Kuyper, H., Buunk, A. P., Van Der Werf, G., & Van Der Zee, Y. (2008). Social comparison in the classroom: A review. Review of Educational Research, 78(4), 828–879. doi:10.3102/ 0034654308321210 Dweck, C. S. (1999). Self-theories: Their role in motivation, personality, and development. Philadelphia, PA: Taylor & Francis. Earley, P. C., & Kanfer, R. (1985). The influence of component participation and role models on goal acceptance, goal satisfaction, and performance. Organizational Behavior and Human Decision Processes, 36, 378–390. doi:10.1016/0749-5978(85)90006-8 8 C´eline Bag`es and Delphine Martinot Else-Quest, N. M., Hyde, J. S., & Linn, M. C. (2010). Cross-national patterns of gender differences in mathematics: A meta-analysis. Psychological Bulletin, 136, 103–127. doi:10.1037/a0018053 French Department of Education. (2007). Diagnostic Assessments – Statistics. Retrieved from http://evace26.education.gouv.fr. Gibson, D. E. (2004). Role models in career development: New directions for theory and research. Journal of Vocational Behavior, 65, 134–156. doi:10.1016/50001-8791(03)00051-4 Good, C., Aronson, J., & Inzlicht, M. (2003). Improving adolescents’ standardized test performance: An intervention to reduce the effects of stereotype threat. Applied Developmental Psychology, 24, 645–662. doi:10.1016/j.appdev.2003.09.002 Huguet, P., & R´egner, I. (2007). Stereotype threat among schoolgirls in quasi-ordinary classroom circumstances. Journal of Educational Psychology, 99, 545–560. doi:10.1037/0022-0663.99. 3.545 Javidan, M., Bemmels, B., Devine, K. S., & Dastmalchian, A. (1995). Superior and subordinate gender and the acceptance of superiors as role models. Human Relations, 48, 1271–1284. doi:10.1177/001872679504801102 Keller, J., & Dauenheimer D. (2003). Stereotype threat in the classroom: Dejection mediates the disrupting threat effect on women’s math performance. Personality and Social Psychology Bulletin, 29, 371–381. doi:10.1177/0146167202250218 Lockwood, P. (2006). Someone like me can be successful: Do college students need same gender role models? Psychology of Women Quaterly, 30, 36–46. doi:10.1111/j.1471-6402. 2006.00260.x Lockwood, P., Jordan, C. H., & Kunda, Z. (2002). Motivation by positive or negative role models: Regulatory focus determines who will best inspire us. Journal of Personality and Social Psychology, 83, 854–864. doi:10.1037//0022-3514.83.4.854 Lockwood, P., & Kunda, Z. (1997). Superstars and me: Predicting the impact of role models on the self. Journal of Personality and Social Psychology, 73, 91–103. Marx, D. M., & Roman, J. S. (2002). Female role models: Protecting women’s math test performance. Personality and Social Psychology Bulletin, 28, 1183–1193. doi:10.1177/ 01461672022812004 Mueller, C. M., & Dweck, C. S. (1998). Praise for intelligence can undermine children’s motivation and performance. Journal of Personality and Social Psychology, 75, 33–52. Nye, B., Konstantopoulos, S., & Hedges, L. V. (2004). How large are teacher effects? Educational Evaluation and Policy Analysis, 26, 237–257. doi:10.3102/01623737026003237 Steele, J. (2003). Children’s gender stereotypes about math: The role of stereotype stratification. Journal of Applied Social Psychology, 33, 2587–2606. doi:10.1111/j.1559-1816.2003. 6602782.X Thoman, D. B., White, P. H., Yamawaki, N., & Koishi, H. (2008). Variations of gender-math stereotype content affect women’s vulnerability to stereotype threat. Sex Roles, 58, 702–712. doi:10.1007/s11199-008-9390.x Van Yperen, N. W., Brenninkmeijer, V., & Buunk, A. P. (2006). People’s responses to upward and downward social comparisons: The role of the individual’s effort-performance expectancy. British Journal of Social Psychology, 45, 519–533. doi:10.1348/014466605X53479 Vezeau, C., Chouinard, R., Bouffard, T., & Couture, N. (1998). Adaptation et validation des e´chelles de Fennema et Sherman sur les attitudes en math´ematique des e´l`eves du secondaire [Adaptation and validation of the Fennema Sherman Mathematics Attitudes Scales in mathematics throughout childhood and adolescence]. Canadian Journal of Behavioral Science, 30, 137– 140. doi:10.1016/50008-400X(02)00112-4 Weiner, B. (1985). An attributional theory of achievement motivation and emotion. Psychological Review, 92, 548–573. Received 24 February 2010; revised version received 8 December 2010

© Copyright 2018