Title bar


Stereotype Vulnerability: Effects on Test Performance

Persistent differences in performance across racial/ethnic groups, even among populations comparable on many factors known to be correlated with achievement, have been the focus of research seeking to determine underlying causes. One barrier that may have been underappreciated is negative stereotypes. Efforts to establish the effects of stereotypes have yielded insights into institutional factors or situations that influence the measurement of educational outcomes, such as standardized test performance.
The concept of "stereotype vulnerability" has been applied to testing conditions affecting black students, though the phenomenon is not restricted to members of a single group. [11] A stereotype can place a member of a group in a social-psychological predicament. The mere existence of a culturally held stereotype about a group to which one belongs means than anything one does, or anything about oneself that fits the stereotype, threatens to confirm it as a self-characterization. When the stereotype threatens an important aspect of the self, such as test performance, the vulnerability can be disturbing. The stereotype may set up a pressure that can undermine performance.
As a test of the theory, black and white undergraduates with comparable Scholastic Aptitude Test (SAT) scores were given tests of items selected from the Graduate Record Examination, chosen to present a high level of difficulty to those completing them. Test-taking conditions were carefully controlled, with the experimental manipulation being subtle but explicit differences in key phrases used in the administration of the tests. For example, in the diagnostic condition, subjects read that the main objective of the experiment was to measure ability in order to understand what personal factors are associated with high or low ability. In the nondiagnostic condition, the purpose was stated as examining the psychology of problem solving. A second round of tests was administered to examine aspects of the experimental manipulation: All students were informed that they should expect a low performance level, and were then monitored on such variables as time spent on the test and stress levels produced.
Results from the two experiments showed significant effects related to race and condition. With SAT differences statistically controlled, blacks performed significantly worse than whites when the test was presented as a measure of ability, but at the same level as whites when the test was presented in less evaluative terms.
"Conditions designed to make black subjects stereotype vulnerable depressed their performance relative to white subjects. Conditions designed to alleviate this vulnerability improved blacks' performance, completely equating the two groups.... Thus, in addition to whatever environmental or genetic influences on the skills that a person brings to the testing situation, the present research shows that the situation itself is not neutral-not even when the tester is the same race as the test-taker and shares his or her dialect. It is a microcosm of the individual's predicament in the larger society; the same stereotypes that make them vulnerable there have the power to threaten them and undermine their effectiveness in the testing situation. This is increasingly so as the test includes more items at the limits of, or beyond, the skills of the test-takers, thereby increasing frustration. By providing evidence of this process, the present results raise the possibility that some portion of the black-white difference in standardized test performance is attributable to the biasing effects of stereotype vulnerability in the testing situation" (Steele and Aronson 1994, p. 15).
11. Several studies examining this phenomenon have been conducted under the direction of Claude M. Steele, Department of Psychology, Stanford University. The particular experiment described here was reported in Steele and Aronson 1994. Up arrow
TOC buttonHelp buttonAbbreviations 
buttonNSF buttonLeft button