Fairtest Examiner - Stereotypes Lower Test Scores Fall 1995

New research by a Stanford University professor demolishes claims by testmakers that racial and gender differences in standardized test performance can be explained fully by differences in academic prowess and preparation. Claude Steele, a professor of social psychology and president-elect of the Western Psychological Association, reports that his seven-year study shows that minority and female students who are aware of racial and gender stereotypes about intellectual ability will score lower on those standardized tests that purport to measure academic aptitude. The study defines this extra burden borne by some test-takers as stereotype vulnerability. 

 Consider the black student who gives the wrong answer or falters grammatically in class, explains Steele. He is vulnerable to the judgment, as is anyone, that he lacks a particular skill.  But he is also vulnerable to the deeper devaluation contained in the stereotype that he has confirmed. Likewise for women in math and sciences. 

In more than a dozen experiments over the past four years, Steele and his colleagues were able to depress the performance of high-achieving African American men and women of all races by subtly implying that well-known stereotypes about those groups intellectual abilities might apply to the test they were about to take. The cues were often indirect. Students were told that the test they were about to take can measure ability or they were asked to mark down their races before the test began. In control groups where similar students were given no reason to suspect that the demeaning stereotypes would apply to their performance, African Americans performed as well as whites on very challenging tests. 

 Even students who do not consciously embrace the stereotypes experience vulnerability. Steele found that minority students often redouble their efforts during a test in order to disprove stereotypes but then end up working too quickly or inefficiently. He explained that performing in domains where prevailing stereotypes indicate that one may be part of an inferior group carries the risk that any faltering of performance will confirm the stereotype as a self-characteristic. 

This dynamic, Steele added, may lead students from stereotyped groups to alternate between trying to do the scholastic task and thinking about what their performance means. Their performance can be disrupted by interfering anxiety, reticence to respond and distracting thoughts. The stakes will be higher for African American and other minority students because a poor performance on a standardized test has a more devastating meaning for these students. 

Steele found that everyone is susceptible to stereotype vulnerability: in one experiment, white males unaccustomed to being intellectually stigmatized were told that Asians achieved higher scores than Americans on a mathematics test they were asked to take. This group achieved lower scores than a control group of white males who were not told anything about previous test results. Steele and a colleague, Steve Spencer at Hope College in Michigan, have also run eight experiments showing that stereotype vulnerability can negatively affect women who believe a given math test shows gender differences. As numerous studies have demonstrated, standardized tests such as the SAT consistently underpredict the performance of women in college. The stereotype vulnerability research suggests that the inherent gender bias of the SAT is made worse by female test-takers awareness of that bias. 

Professor Steele noted that his findings underscore the danger of relying too heavily on standardized test results in college admissions or otherwise. The research findings make very clear the danger of all practices including the National Merit Scholarship Semifinalist selection process and the NCAAs initial athletic eligibility rules that base high stakes decisions on standardized test scores. 
Claude Steele has Scores to Settle NY Times Magazine 17 Sept 1995 
Why do African-Americans score significantly worse on standardized tests than whites? 
"Our idea," Steele says to the group, "was that whenever black students concentrate on an explicitly scholastic task, they risk confirming their group's negative stereotype. This extra burden, in situations with certain characteristics, can be enough to drag down their performance. We call this burden stereotype vulnerability." 

In the first experiment Steele describes, he and Joshua Aronson from the University of Texas gave two groups of black and white Stanford undergraduates a test composed of the most difficult verbal-skills questions from the Graduate Record Exam. Before the test, one group was told that the purpose of the exercise was only to research "psychological factors involved in solving verbal problems," while the other group was told that the exam was "a genuine test of your verbal abilities and limitations." 

"This is what we found," Steele says, placing a transparency onto the overhead projector. As the information in the bar charts sinks in, people sit up in their chairs. There are several audible "hmmms," a muffled "wow!" Then a professor at the back of the room asks: "Did you give the groups the same test?" 

Steele smiles and says, "Yes." The question speaks to the startling nature of the results. As the graphs indicate, the blacks who thought they were simply solving problems performed as well as the whites (who performed equally in both situations). However, the group of black students who labored under the belief that the test could measure their intellectual potential performed significantly worse than all the other students. 

Steele's idea of stereotype vulnerability is not that the student consciously or unconsciously accepts the stereotype (as other social scientists have speculated), but rather, as Steele says, that they have to contend with this whisper of inferiority at the moment when their mental abilities are most taxed. In trying not to give credence to the stereotype, Steele theorizes, the students may redouble their efforts only to work too quickly or inefficiently. The cues that can spark the vulnerability can be subtle-like suggesting the test can measure ability or making students mark down their races before the test begins. While there might be no perceptible bias in a given test or in the test-taking situation, an exam might still be weighted against blacks because the possibility of performing badly has a more devastating meaning. 

As Steele goes on to describe his experiments-more than a dozen over the last four years-the audience remains riveted. With his colleague Steve Spencer at Hope College, he has run eight additional experiments showing that stereotype vulnerability can negatively affect women who believe a given math test shows "gender differences." The negative impact of stereotype vulnerability has even appeared in white men who took a difficult math test after being told that Asians tended to outperform whites on that particular exam. 

At Michigan, Steele found a puzzle in the statistics on the performance of black students that drew his interest. According to university data, no matter how well qualified a black student was at a given level, he or she was more likely than an equally qualified white student to fall behind at the next level of achievement. Steele figured that the more well-known factors like discrimination, bad schools, broken homes and the poverty and crime that disproportionately damage black communities couldn't fully account for the plight of highly motivated, middle-class African-Americans. "Black students with 1,300 S.A.T. scores were going home from college with 2.4 G.P.A.'s," Steele says. "When I realized that the smartest black students were having these terrible troubles, I figured something else was going on." 

After talking with students, Steele decided that the answer was not a lack of a desire to succeed. Nor could he find significant overt or covert racism on the part of the school. He began formulating theories on how the ebbs and flows of anxiety and confidence that all students experience might be exaggerated in black students by the deeper currents in a racially focused society. 

As Steele describes it, stereotype vulnerability is a "patient predator" that affects black students not just in testing situations, but in every area of their academic lives. "At each new proving ground the problem can re-emerge," Steele says. "The student who has risen above the stereotype in high school-proving to everyone that he is smart-goes off to college, where a new level of performance is expected." 

Eventually, even minor failures-a lapse in grammar while talking to a professor-can discourage. At some point, students may begin to "disidentify" or pull away from the challenge of school. 

But Steele doesn't just want to explain away the overall poor showing of black students. "One of the thrills of doing this research," Steele says, "is that I can provide an optimistic take without being political-it's in the results." The results he's referring to are not the ones that demonstrate that stereotype vulnerability exists, but those that show that it disappears with the subtlest of changes. Take away the situational clues that make the setting racially charged (like having to identify race before the exam), and the vulnerability can vanish from the student's performance. If this vulnerability can be consistently kept at bay, Steele says, the downward spiral may never start. 

Steele says that those who have seized on the apparent willful pulling away of the black student have unfairly focused on the very end of a long descent that can begin with a vulnerability so subtle that not even the student can identify its origin. 

"These kids want to make it," Steele says. "These kids are in their dorm rooms confronting the beast of their deepest fears. Instead of telling them that it is their responsibility to overcome the psychology that they are feeling in their gut, let's get rid of the beast." 

"I think much of what is mistaken for racial animosity in America today is really stereotype vulnerability," he says. "Imagine a black and a white man meeting for the first time. Because the black person knows the stereotypes of his group, he attempts to deflect those negative traits, finding ways of trying to communicate, in effect, 'Don't think of me as incompetent.' The white, for his part, is busy deflecting the stereotypes of his group: 'Don't think of me as a racist.' Every action becomes loaded with the potential of confirming the stereotype, and you end up with two people struggling with these phantoms they're only half aware of. The discomfort and tension is often mistaken for racial animosity." 

The parallel is instructive. The white person who has felt how much energy it takes to communicate the message "I am not the racist person you might think I am" can perhaps also understand how distracting such an effort might be during a standardized test. Following the idea of white vulnerability, he says he can even explain something of the current frustration with racial issues. 

"Many whites are just like the black kid in the school situation," Steele says. "When racial issues become frustrating-when there are setbacks -- the suspicions about their sincerity re-emerges. If they feel that, in continuing to try, they will only give more evidence to the conclusion that they are really racist they'll eventually want to withdraw. That is what is happening: the whole thing has become too loaded, and we're seeing whites walking away.