Creating, Closing, and Reversing the Gender Gap in Test Performance

Abstract

We investigate how selection policies—the rules defining access to a valued position—can act as situational cues signaling social identity threat or safety among women and men. College students took a logic test ostensibly determining their assignment to a position of leader or subordinate for a subsequent task. Study 1 showed that when only the test score determined the selection, women experienced more identity threat and performed worse than men. When the policy allowed the selection of women at a lower level of performance than men to promote diversity, men’s performance decreased compared to the merit condition, falling to the level of women’s performance and thus closing the gender gap. Study 2 replicated these findings and established that the meaning derived from selection practices affects candidates’ performance. A third policy that also preferentially selected women, but to correct for unequal treatment based on gender, leads to a reversed gender gap (i.e., women outperformed men). These findings suggest that structural features of test settings including selection practices can constrain individuals’ potential access to opportunities.