Stereotype threat among schoolgirls in quasi-ordinary classroom circumstances
Two experiments investigated whether school children are influenced by stereotype threat in regular classroom environments. In Experiment 1, highly math-identified boys and girls between 11-13 years in age were presented with a version of the Rey-Osterrieth Complex Figure (ROCF) recall memory task, an exercise believed to tap a variety of skills essential for successful academic performance. Half the students were told that the task was a "geometry test" (stereotype threat for girls) and the other students were told it was a "memory game" (no stereotype threat). Results showed that girls' memory performance was superior when the task was described as a memory game compared with a geometry test, and boys' performance was better when the task was described as a geometry test compared with a memory game. Experiment 2 provided a conceptual replication, but in an ordinary classroom environment. Experimenters entered classrooms and created two mixed-gender or same-gender groups of students. Students were presented with the ROCF and were told either that the task provided a test of their "ability in geometry" or their "ability in drawing." After entering math grades as a covariate, the memory results showed that the results of Experiment 1 replicated in mixed-gender groups. Girls outperformed boys in the "drawing" condition but performed more poorly in the "geometry" condition. In single-sex groups, there were no significant effects of task description. Girls in the "geometry" conditions reported the task as being more difficult than did boys. In addition, when asked to identify students who were high performing in math, girls in the mixed-gender, geometry condition were more likely to nominate a boy than a girl, but girls in the same-gender condition were more likely to specify a girl. The gender of the specified high-performing student mediated girls' memory performance, suggesting that imagining female role-models helped girls to perform well in single-sex settings even under conditions where the task description typically produces stereotype threat. These experiments provide evidence of stereotype threat effects in realistic classroom environments and also highlight conditions that qualify the emergence of performance decrements under threat.