BOX 2-4

Stereotype Threat

In 1995, Claude Steele and Josh Aronson published an influential article in which they demonstrated a phenomenon they called stereotype threat.a Stereotype threat occurs when people feel that they might be judged in terms of a negative stereotype or that they might do something that might inadvertently confirm a stereotype of their group.

When any of us find ourselves in a difficult performance situation, especially one that has time pressure involved, we might recognize that if we do poorly, others could think badly about our own individual abilities. But if you are a woman or minority-group student trying to excel in science or engineering, there is the added worry that poor performance could be taken as confirmation that group stereotypes are valid.

Stereotype threat has been shown to apply to women performing a difficult mathematics test. Women tend to do more poorly than men, not on the average questions, but only on the high-level questions and only when their gender has been commented upon.b When stereotype threat is at work, fewer women will have high scores, and their scores will under-predict their achievement.

A series of studies by Toni Schmader and colleagues suggests that women’s performance can be improved by acknowledging stereotype threat, as shown in Figure B2-4. In one condition, one group of men and women was given a set of word problems and told that it was a problem-solving exercise, with no mention of a test, mathematics, or ability. In this condition (“Problem Solving”), women’s performance on the test was not different from that of their male peers, regardless of whether differences in SAT were controlled for. In a second condition, a different group of men and women was given the same set of word problems and told that their task would yield a diagnostic measure of mathematics ability that would be used to compare men’s and women’s scores; in this condition (“Math Test”), there was a gender gap similar to that seen in SAT-M scores.

In a third condition, a third group of men and women was told that the test they were taking—the same set of word problems as used in condition one and two—was a diagnostic measure of mathematics ability, and that their performance would be used to compare men’s and women’s scores. These are the same conditions that led to performance decrements in the second group. However, they were also informed about stereotype threat and reminded that if they were feeling anxious while taking the test, it might be a result of external stereotypes and not a

fers to the “experience of being in a situation where one faces judgment based on societal stereotypes about one’s group” (Box 2-4).77 For example, women perform worse than men on difficult but not easy math tests if gender stereotypes are made salient or if they are told that the tests have sex differences in performance. But, when women are told that there are no sex


SJ Spencer, CM Steele, and DM Quinn (1999). Stereotype threat and women’s math performance. Journal of Experimental and Social Psychology 35:4-28.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement