When Girls Evaluate Themselves Better than Boys in Minority Groups: Role of the Performance Context

In the literature, token status is more threatening for women than for men. We suggest that the impact of minority status on self-evaluations of boys and girls might depend on the performance context. To test this hypothesis, pupils in the majority or minority by gender took an academic test either in a context favorable to boys (a working-group condition) or favorable to girls (an intergender-comparison condition). As expected, in the working-group condition, girls reported lower self-evaluations than boys did. In the intergender-comparison condition, girls reported better self-evaluations than boys did. The role of gender stereotypes is discussed. Current Research in Social Psychology (Vol. 13, No. 20) Martinot, Désert, & Redersdorff 256 Group gender composition has been shown to affect the effectiveness of women in groups. The fewer women in a group, the less likely their ideas will be considered (Craig & Sherif, 1986). Many laboratory and field studies, in academic as well as occupational contexts, have shown that working in a predominantly male environment creates major difficulties for women (e.g., Alexander & Thoits, 1985; Kanter, 1977; Morrison & von Glinow, 1990; Sackett, DuBois, & Noe, 1991; Spangler, Gordon, & Pipkin, 1978; Yoder, 2002; Yoder & Sinnett, 1985). Accordingly, it has been observed that women's performance suffers if they are the only representative of their gender (Sekaquaptewa & Thompson, 2003) or if their gender is in the minority (Inzlicht & Ben-Zeev, 2003), whether the task to be accomplished is associated with an area where women have a poor reputation (Inzlicht & Ben-Zeev, 2000) or do not have such a reputation (Sekaquaptewa & Thompson, 2002). Other studies have shown that when women must perform in a gender-minority situation, their performance expectations (Sekaquaptewa & Thompson, 2003; Stangor, Carr, & Kiang, 1998) and their self-esteem (Krimmel & Gormley, 2003) are lower; they also receive lower ratings from superiors (Sackett et al., 1991), feel more isolated (Kanter, 1977), and perceive more discrimination (Steele, James, & Barnett, 2002). Although less numerous than studies focusing on women, studies on men working in a minority context have shown that men are apparently not negatively affected by a minority status. Evidence from these studies shows that neither men's performance (e.g., Inzlicht & Ben-Zeev, 2000; Sekaquaptewa & Thompson, 2003), nor their self-perceptions (e.g., Cohen & Swim, 1995; Sekaquaptewa & Thompson, 2003; Yoder & Sinnett, 1985), nor even judgments of them made by others (e.g., Fuegen & Biernat, 2002; Sackett et al., 1991) are lowered by their being in the minority. In sum, the above body of research indicates that women seem to be more negatively affected by a numeric minority status than men. For instance, Policewomen are seen to face many disadvantages, whereas male nurses enjoy advantages from being one of the few among female colleagues (Krimmel & Gormley, 2003). To sum up, previous studies focusing directly, in a same investigation, on the effects of minority status on men's and women's self-evaluations have shown that women's self-evaluations were more threatened than those of men in a context of performance (Craig & Sherif, 1986; Crocker & McGraw, 1984; Krimmel & Gormley, 2003). The aim of the present study was to examine whether the impact of minority status on boys’ and girls’ self-evaluations depends on the performance context. Status and Differential Expectations In their expectation states theory, Berger and his colleagues (Berger, Rosenholtz, & Zelditch, 1980; Berger, Wagner, & Zelditch, 1985), proposed that because gender has traditionally been correlated with prestige and status differences in society, differential expectations exist concerning the social power of men and women. These expectations, in turn, can generalize and, through a process of behavioral confirmation of expectancies similar to the self-fulfilling prophecy (Rosenthal & Jacobson, 1968), affect power-related behavior and perceptions of power across a variety of social contexts (Berger et al., 1985; Eagly, 1983; Meeker & Weitzel-O'Neill, 1985). In line with the expectation states theory, Dovidio, Brown, Heltman, Ellyson and Keating (1988) found that gender differences in power-related behavior were affected by context. They showed that systematic differences in the power-related behaviors of men and women emerged in the situations in which there was differential familiarity based on the gender-linked nature of the task. On the masculine task men displayed more verbal and nonverbal power-related behavior than did women. On the feminine task women exhibited more power than men on most of the verbal and nonverbal measures. In other words, situational cues such as stereotypically male or female tasks can influence behavior because they evoke differential expectations of Current Research in Social Psychology (Vol. 13, No. 20) Martinot, Désert, & Redersdorff 257 expertise, and hence situation-related power or status (Dovidio, Ellyson, Keating, Heltman, & Brown, 1988; Ridgeway, 2001). In a mixed-gender stereotypically masculine work context, because of the effects of gender status beliefs on performance expectations, expectation states theory predicts that men will be more influential than women who are otherwise similar to them (Wagner & Berger, 1997). Furthermore, Williams (1992) argues that men take their gender privilege with them even when they enter predominantly female occupations. This advantaged social status translates into an advantage in spite of their numerical rarity (Ott, 1989; Thompson & Sekaquaptewa, 2002), and allows a “glass escalator effect” (Williams, 1992). Women experience the opposite in maledominated occupations. Many women encounter a “glass ceiling” in their efforts to scale organizational and professional hierarchies (e.g., Williams, 1992). We suggest that, when males work in predominantly female professions (Krimmel & Gormley, 2003; Ott, 1989; Williams, 1992) or when they represent the minority in a working group (Craig & Sherif, 1986), their advantaged social status is likely to be related to a male-favoring gender stereotype. On the contrary, when females work in predominantly male professions (Krimmel & Gormley, 2003; Ott, 1989; Williams, 1992) or when they represent the minority in a working group (Craig & Sherif, 1986), they would not benefit from a female-favoring gender stereotype. If this threat is due to the fact that the performance context makes a male-favoring gender stereotype salient, then women's self-evaluations should no longer be negatively affected when a gender stereotype favorable to women is made salient. Indeed, as expectation states theory argues, the biasing effect of gender status on self-other performance expectations is not invariant across all situations. Although they are primed by gender categorization in all situations (Fiske, 1998), the diagnostic value of gender status beliefs to participants can vary from very little to substantial, depending on how salient and task relevant gender is the situation compared to other social roles or status-valued distinctions that are also salient for the participants (Wagner & Berger, 1997; Ridgeway, 2001). There are other examples in the literature of situations in which the performance context has a beneficial impact on women, even when they have to perform on a domain related to a negative reputation of their group. In line with stereotype threat research (e.g., Shih, Pittinsky, & Ambady, 1999; Spencer, Steele, & Quinn, 1998), different stereotypes triggered by the performance context can be activated. For example, in the United States, Shih and colleagues (e.g., Shih et al., 1999) showed that Asian American women did better on a mathematics test when their Asiatic identity was made salient than when their identity as females was brought to the fore. These authors account for these results in terms of the activation of identity stereotypes, here, the stereotype that Asiatics excel in mathematics as opposed to the stereotype that women's aptitude is poor in this domain. In line with such results, we believe that self-evaluations of boys and girls may depend on the activation of different gender stereotypes triggered by the performance context and the participant's minority status. The main objective of our experiment was to test this hypothesis. Minority Status and Gender Stereotypes To be in the minority is to be put under a magnifying glass, accentuating the differences between majority and minority (e.g., Kanter, 1977). Moreover, a phenomenon of assimilation causes members of the minority to be judged, and to judge themselves, in a more extreme and stereotyped manner than do members of the majority group (Cohen & Swim, 1995; Crocker & McGraw, 1984; Kanter, 1977). Thus, males and females are seen as more likely to be playing out Current Research in Social Psychology (Vol. 13, No. 20) Martinot, Désert, & Redersdorff 258 a gender-stereotyped role, the fewer the number of other members of their gender subgroup are present. When they are in the minority in a working group, men are likely to be perceived as more influential and leaders within the group than women (Craig & Sherif, 1986; Crocker & McGraw, 1984; Jemmott & Gonzales, 1989; Williams, 1992). Indeed, expectation states theory argues that gender is deeply entwined with social hierarchy and leadership because the rules for the gender system that are encoded in gender stereotypes contain status beliefs at their core (Wagner & Berger, 1997). Perceptions of leadership skills among men and influence exercised over other group members are positively correlated with the proportion of women present in the performance context (Ridgeway & Balkwell, 1997; Tsui, Porter, & Egan, 2002; Webster & Hysom, 1998). Social norms are such that men occupy higher and more prestigious positions than women (Cotter, DeFiore, Hermsen, Kowalewski, & Vanneman, 1997); men are also perceived as better suited to the role of leader (Crocker & McGraw, 1984; Powell, Butterfield, & Parent, 2002). Thus, the greater the ratio of women to men in the performance context, the more men are perceived as having good leadership skills. In other words, men are evaluated favorably when they are in the minority because a positive stereotype is activated (Fuegen & Biernat, 2002). On the other hand, because the content of the female stereotype does not involve leadership skills (Bem, 1974; Williams, Daws, Best, Tilquin, Wesley, & Bjerke, 1979), women are more likely to be perceived as less influential and as having weaker leadership skills in a working group where they are in the minority (Craig & Sherif, 1986). As a consequence, we suggest that the impact of minority status on self-evaluations should be more negative for women than for men when they must perform in a working group. Nevertheless, we suggest that, in a performance context that makes a female-favoring gender stereotype salient, this effect on self-evaluations should be reversed. Recently, Martinot and Désert (2007) showed that French middle-school pupils were not only aware of, but also endorse, a gender stereotype regarding academic abilities that favors girls over boys. French middleschool boys and girls think that people in general perceive girls as being more academically able than boys, and they also believe this personally. This stereotype is particularly interesting because it relates to a socially important domain -academic success -and because it clearly favors girls. The fact that the French academic context conveys this stereotype makes it an ideal setting for studying the effects of numerical asymmetries in a domain that is both socially valued and favorable to girls. Moreover, it is interesting to test if the effects observed among adults can be replicated among pupils. Accordingly, we suggest that a girl-boy comparison on an academic test constitutes a performance situation that activates a gender stereotype favoring girls. For this reason, we hypothesize that, in a setting that induces an intergender comparison of academic abilities among participants in the minority by gender, self-evaluations of girls would be higher than those of boys. On the other hand, to solve problems of an academic test in a working group constitutes a performance situation that activates the gender stereotype bearing on boy's superior leadership skills. Therefore, in the “working group” condition, when pupils are placed in the minority by gender, girls should report lower self-evaluations than boys. Furthermore, since belonging to the majority in a performance context makes individuals less visible and less distinct (Crocker & McGraw, 1984; Kanter, 1977), gender identity and its associated stereotypes should be much less salient among pupils who are in the majority based on gender. Hence, the self-evaluation differences between girls and boys should be reduced in this case, whatever the performance context (intergender comparison or working group). Current Research in Social Psychology (Vol. 13, No. 20) Martinot, Désert, & Redersdorff