Activating stereotypes undermines task performance expectations
Two experiments examined the consequences of stereotype activation on expectations of performance. Students in these experiments expected to complete two tasks but actually completed only one. After students completed the first task involving finding words in an array of letters, they were told that they did either well or relatively poorly on the task. They then were told that they would complete a second test but, prior to doing so, were asked to estimate how well they would perform. In Experiment 1, participants were female students and the second task supposedly involved spatial abilities. The description of the task was systematically varied, such that half the participants were told that men outperform women on the task (stereotype threat) and half were told that there were no gender differences on the task (control). Students then estimated how well they expected to do on the upcoming task. In the control condition, women were more optimistic about their performance on the second task if they had received positive (compared with negative) feedback on the first task. In the stereotype threat condition, however, women showed low expected levels of performance regardless of feedback they had received after the first task. In Experiment 2, male and female students were asked to estimate how well they would perform on the second task if they performed in a group of similar or dissimilar students. Prior feedback predicted performance expectations when students thought they would work with similar others but was unrelated to expectations when students thought they would work with individuals who were different from themselves. Stereotype activation and stereotype threat appeared to undercut the gains in confidence and expectations of success that are typically provided by positive feedback.