Science Directions

Every now and then, one is confronted with a research presentation that is so powerful and so far-reaching in its implications that it begs for the widest possible dissemination. Such was the feeling I had when I attended the Multicultural Conference and Summit.

Indeed, I didn't have to wait very long after the conference started to experience this effect. The first speaker was Professor Claude Steele of Stanford University and the results he presented were unbelievably powerful to me. Findings from his laboratory and the work of others have provided us with insights into underachievement tied to the activation of negative stereotypes. These effects are consistent across many different groups, and remarkably easy to elicit.

How many times have you heard the statement that African-American and other ethnic-minority students score significantly lower on average on the SAT and the Graduate Record Examination (GRE) compared with white students? For many years, sub-par SAT and GRE scores were used to justify the rejection of ethnic-minority applicants by undergraduate and graduate programs in many colleges and universities. The assumption was that scores on these two standardized tests provided a measure of intellectual ability, and ethnic-minority applicants often did not measure up. Questions arose about the cultural bias of standardized tests and some undergraduate and graduate admissions officers examined the total record of a given applicant and not just that person's SAT or GRE scores. Well, what if the assumption about cultural bias was not the critical issue? What if some ethnic-minority students performed less well on average on standardized tests because of subtle cues that activated the negative stereotype about ethnic-minority intellectual abilities that has become so ingrained in our society? That is the crux of Steele's research program.

Consider the following experiment from Steele's laboratory: African-American and white undergraduates are brought into the laboratory and presented with a series of questions taken from the SAT. One group of African-American and white students is told that the questions they will answer provide a sensitive measure of intellectual ability. The other group of African-American and white students is only instructed to answer the questions. The scores of white students in the two groups did not differ. However, the scores of African-African students in the group that was told that the questions provided a sensitive measure of intellectual ability were significantly lower than the scores of African-American students who were not provided with this information. Importantly, the African-American students who did not receive the information about intellectual measures performed just as well as the white students. What an eye-opening finding. Rather than a cultural bias in standardized tests, Steele's research suggests a powerful influence of negative stereotypes on the performance of African-American students. Societal expectations of lower test scores by African-American students are realized only if the negative stereotype is activated during the actual testing.

Is this effect limited to African-American students taking standardized tests? Quite the contrary! Two groups of white engineering students are presented with difficult math problems. One group is encouraged to do the best they possibly can. The other group is told that Asian-American engineering students are especially adept at solving these math problems. The latter group performed significantly worse than the former group, apparently because of the activation of the negative stereotype that Asian-American students are superior to white students in math ability.

Additional research on negative stereotypes has been linked to gender differences in math scores and in racial differences in athletic ability. Taken together, these studies appear to tap into the cumulative effects of these pervasive negative messages that have saturated our society for decades. Obviously there are individual differences in sensitivity to the effects of negative stereotypes on performance. However, there is little doubt that many people are adversely affected, especially within the educational system.

A next step in this area of research will be to demonstrate these effects in real-world settings. If these findings stand the test of time, intervention strategies might be developed to buffer children and adults against the deleterious effects of negative stereotypes. In addition, broader educational policies and practices should consider carefully the implication of these findings for teacher training and curriculum development.