Stereotype threat: The effect of expectancy on performance

The goal of this study was to investigate to role of expectancy as a potential mediator of performance deficits under stereotype threat. In Experiment 1, female students were assigned to one of three experimental conditions in which they were told that women perform worse (Negative information), equally (Control) or better (Positive information) than men in logical-mathematical tests. Later, they were given a difficult math test and asked to estimate their performance prior to taking the test. Consistent with predictions, participants who considered logical-mathematical abilities important and received negative information regarding the ingroup showed lower levels of expectations and a sharp decrease in performance compared to women in the positive and control conditions. Moreover, expectancy was found to partially mediate the effect of stereotype threat on performance. In Experiment 2, we tested the generalizability of these results to non-stigmatized groups. A group of Black Americans living in Italy were provided with favorable or unfavorable information about either their minority (Blacks) or their majority (Americans) ingroup. Consistent with predictions, participants both in the minority and in the majority condition had lower expectations and under-performed after negative information about the ingroup. However, the level of expectancy was found to mediate the decrease in performance for participants in the Black but not in the American condition. In the discussion of these results it is suggested that, although comparable performance deficits are found for minority and majority members, the underlying processes may be different.