Gender Identification Moderates Stereotype Threat Effects on Women's Math Performance
This study examined the moderating role of group identification on stereotype threat. Steele & Aronson (1995) and others had shown that individuals are more likely to show performance decrements under conditions creating stereotype threat when they strongly identify with the domain in question (e.g., women who strongly identified with mathematics). This study sought to test whether the strength of one's group identification might also play a role (e.g., women whose gender is central to their self-definition). To do so, men and women college students who varied in terms of the centrality of gender to their identity (based on their responses to a questionnaire given earlier in the semester) were recruited to answer items that were to be included in a new math test. All students were told that the researchers would be comparing each student's score against the scores of the other students. For half of the students, however, the link between their gender and their performance was highlighted. These students were told that a comparison between the performance of men and women was central to the study and that each individual's score would be used to predict the performance of men and women more generally (stereotype threat for women). In the other condition, no mention of gender in relation to performance was made (control). Results indicated that for women in the stereotype threat condition, those whose identity emphasized gender performed more poorly than men, and women whose gender was less central to their identity performed equivalently with men. In the control condition, however, women performed equally with men and there were no effects of gender identity. This study shows that stereotype threat effects are most common when one's identification with both the content domain and with the relevant group is highlighted in a social context.