ReducingStereotypeThreat.org

Keller, 2007a

This study involved a manipulation of stereotype threat, measures of domain identification, and test items of varying difficulty. Although these factors have been examined in separate settings, this study examined the effects of all three variables within a single design and in a realistic school setting. Male and female German secondary school students who took math as a regular subject were asked to complete a difficult math task under one of two conditions. Half of the students were told that the task tends to produce gender differences (stereotype threat for girls), whereas the other half of the students were told that there were no gender differences on the task (control). Under stereotype threat, girls who highly identified with math performed more poorly than girls not identified with math. In the control condition, in contrast, highly-identified girls outperformed low-identified girls. Boys were not affected by the stereotype threat manipulation. Analyses including item difficulty showed that low-identified girls performed better on difficult items under stereotype threat than under control, but highly-identified girls performed more poorly on difficult than easy items under stereotype threat. These results show that threat heightens motivation with individuals low in identification producing enhanced performance, but stereotype threat undermines performance of individuals who are highly-identified with a domain, particularly on difficult test items.