Abstract
Previous research has found that activating negative stereotypes after completion of a task can lead people to feel more certain that they performed poorly (i.e., stereotype validation). The current research examined the implications that stereotype validation may hold for future performance. In two studies, the accessibility of gender stereotypes was manipulated after participants completed an initial test on a topic. After reporting perceptions of their performance, participants completed a follow-up test on the same subject. Consistent with hypotheses, activating negative stereotypes after the initial test was found to increase certainty in negative performance evaluations among stigmatized participants. Importantly, this higher certainty – triggered by stereotypes – predicted decreased performance on the later, follow-up test.
Notes
1. The Gender × Gender Stereotype interaction was significant when perceived and actual performance on the first quiz were not controlled via a two-way ANOVA, F(1, 367) = 5.31, p = .022, = .014.
2. In addition, a supplementary centered regression analysis revealed that the Gender × Gender Stereotype interaction was not qualified by differences in perceived performance (controlling for actual performance), Perceived Performance × Gender × Gender Stereotype, b = .04, t(362) = .24, p = .810.
3. A two-way ANCOVA (controlling for pre-manipulation actual and perceived performance) was also conducted on the full sample that included participants who failed the attention check. The Gender × Gender Stereotype interaction was found to be directionally consistent with predictions, but not statistically significant, F(1, 437) = 2.92, p = .088, = .007.
4. Questions were adapted from content on the following websites: http://www.indiabix.com/general-knowledge/general-science/, http://time.com/3394642/quiz-5th-grade-science/, http://time.com/9571/basic-science-quiz/, http://www.indiabix.com/general-knowledge/physics/010002, and http://www.glencoe.com/qe/scienceOLC.php.
5. A three-way ANCOVA revealed a significant three-way interaction, F(1, 562) = 6.32, p = .012, = .011. For participants without a postgraduate degree, the stereotype manipulation influenced the certainty of women more than men – with women reporting greater certainty in poor performance in the stereotype-present compared to stereotype-absent condition. Among participants with a postgraduate degree, men were not influenced by the manipulation. However, women with a postgraduate degree were more certain they had performed poorly when the stereotype was absent rather than present.
6. A two-way ANOVA showed an interaction pattern that was directionally consistent with predictions (when pre-manipulation actual and perceived performance were not controlled), but this effect was not significant, F(1, 548) = 2.05, p = .153.
7. A centered regression found that the Gender × Gender Stereotype effect was not qualified by differences in perceived performance (controlling for actual performance), Perceived Performance × Gender × Gender Stereotype, b = .20, t(543) = 1.68, p = .094.
8. When including all participants who failed the attention check, the Gender × Gender Stereotype interaction was found to be significant via a two-way ANCOVA (which controlled for pre-manipulation actual and perceived performance), F(1, 627) = 6.65, p = .010, = .010.