Are Stereotypes True?
In an extensive review, Steele (1997) indicated several factors that need to be present for stereotype threat to have negative effects on performance. First, negative stereotypes about a group must exist for the testing domain, as in the example of women and math used above. Other domains include African Americans and intelligence, the elderly and memory, and women and driving, among others. Second, the stereotype must be relevant to the situation. For example, an African American man will be concerned about confirming the stereotype that African Americans are less intelligent than Caucasians while taking the SAT, but not while playing basketball. Thus, stereotype threat may affect his performance while taking the SAT but not while playing basketball. Third, the testing domain must be difficult enough so that poor performance is likely and could be perceived as stereotype confirming. If a memory test is too easy, an elderly person will not experience frustration and doubt, and thus will not be concerned with confirming the stereotype that the elderly have worse memories than younger people. Fourth, stereotype threat is most harmful to people who identify with the domain and consider it important to their self-worth. For instance, women who gain their self-worth from success in math will be most concerned with being viewed as confirming the negative stereotype. Women who do not gain their self-worth from math will not be as concerned about performing poorly in that domain. As another example, Danica Patrick, a NASCAR racer, would be more concerned about performing poorly in the domain of driving and confirming the stereotype that women can’t drive than Mary Lou Retton, a famous gymnast. Finally, the stereotype must be prevalent and widely known. If a person does not know about the stereotype, they will not be concerned with confirming the stereotype. When the above factors are present, stereotype threat and its negative consequences are likely to ensue.