The effect of negative performance stereotypes on learning
This paper investigated the effects of stereotype threat on learning in three experiments. Experiment 1 involved female undergraduate students (N = 59) who were taught a novel math task through a two-part tutorial, each revealing half of the rules necessary for the learning test. Between the two parts, participants were given instructions either with no reference to gender (control condition) or stating that the purpose of the study was to examine why women were worse than men at math (stereotype threat). After the tutorial, participants completed math problems and a learning test. Performance did not differ for material presented in the first half of the tutorial, but women in the stereotype threat condition learned fewer of the rules presented in the second half of the tutorial, after the manipulation of stereotype threat. These findings suggest that stereotype threat reduced womens ability to learn mathematical rules. Experiment 2 was conducted with female undergraduates (N = 92) who attempted to learn a set of mathematical rules when told that the study examines why women were worse than men at math (stereotype threat) or not (control). Importantly, this manipulation occured either at the beginning (stereotype threat prior to learning) or at the end (stereotype threat after learning). Women who received the stereotype threat instructions before learning showed reduced ability to explain the task and performed worse on even the easiest math problems. These effects were not present when the instructions were presented after learning. Performance on the easy problems in the stereotype threat prior to learning condition was predicted by study time, suggesting that stereotype threat impaired womens attention to math-related materials. Experiment 3 was conducted with participants (N = 81, 50 women, 31 men) randomly assigned to control or stereotype threat conditions. Participants learned how to complete a focal task, then completed a transfer task using the same underlying mathematical principles as the focal task, and then an implicit learning task which was based on the rules of the initial focal task. They then completed a measure of threat-based concern. There was no difference in mens threat-based concern between the control and stereotype threat conditions, but women in the stereotype threat condition reported more threat-based concern than those in the control. Mens performance on all of the tasks was consistent between the control and stereotype threat conditions, but women in the stereotype threat condition performed worse on all of the tasks than those in the control, particularly the transfer task. These results indicate that stereotype threat reduces learning, in addition to harming execution of skills in the stereotyped domain.