Earlier this month, a study suggested  that the worry of potentially confirming the “dumb jock” stereotype might be to blame for the performance gap between male athletes and non-athletes at selective liberal arts institutions. Now, another study  suggests that this phenomenon – known as “stereotype threat” – could also be to blame for some of the underperformance of minority and female students on standardized tests and in the college classroom.
Gregory M. Walton and Steven J. Spencer, psychology professors at Stanford University and the University of Waterloo, respectively, recently conducted a meta-analysis of existing data from other studies to test the existence of a stereotype threat for non-Asian minority students and women in quantitative fields. Their research, soon to be published in Psychological Science, confirms their hypothesis that these stereotyped groups actually perform better than non-stereotyped groups at the same level of performance when this threat is removed from the academic environment.
When students are taking standardized tests such as the SAT and GRE, Walton pointed out, this threat is triggered in minute ways, such as asking a test taker to self-identify his or her race and gender prior to taking the test. He noted, however, that a test taker might not even have to be triggered in this explicit way, but might already be hyperaware of a stereotype “impugning the ability of [his or her] ethnic or gender group.”
Using data from the SAT, the study found that “stereotype threat” reduced the scores of women on the math section by 19-21 points. This depressed score is especially significant since the overall gender gap on this section is 34 points. The study also found that “stereotype threat” reduced the scores of African and Hispanic Americans by 39-41 points. The overall gaps between these groups and white students are 199 and 148 points respectively.
“Like the time of a track star running into a stiff headwind, such performances underestimate the true ability of stereotyped students,” the study said of these individuals’ scores on standardized tests.
Walton said test makers could address part of this problem by putting demographic questions about race and gender at the end of the test instead of at the beginning. Those administering the test, he added, could also assist by conducting self-affirming tasks immediately before taking a standardized test. Walton said some studies have shown that allowing students to write briefly about “something that they value or is important to them” can improve performance among stereotyped groups.
Still, he cautioned these simple steps alone would not be enough to reverse “stereotype threat,” as the problem is environmental rather than with the standardized tests themselves.
“There’s no magic bullet,” Walton said. “The problem that we’re talking about is a family of related threats. There’s no one threat and no one solution to that threat. All of these threats differ in subtle ways and different ways. Though there’s no simple solution, there are interventions that show enormous promise. The critical task for future research is to work together with schools and companies to work on these interventions and test them. At the moment, we don’t know how important [they'll be] or how well they’ll work.”
In the classroom, Walton said, stereotype threat can be significantly reduced by teaching potentially stereotyped students about the phenomenon. This, he argued, allows these students to attribute feelings of anxiety or arousal about academics to that threat rather than to a personal or predisposed risk of failure.
He noted that such an experiment in intervention is already taking place at Waterloo – his co-author’s institution, located in Ontario. Some faculty members at the university’s college of engineering are working with first-year engineering students to help them understand how older students experienced their transition to the institution and dealt with stereotype threat.
The next step for researchers, Walton said, is to scale up interventions such as these to “larger and more heterogeneous populations” to judge their effectiveness. Until then, the stereotype threat persists.