Stereotype Threat and Female Students' Math Performance
There is no shortage of stereotypes about the intellectual abilities of social groups, and by a young age, we know them very well. By middle childhood, most American children are familiar with stereotypes portraying boys as better at math than girls, black people as less intelligent than white people, Asians as better at math than anyone else, and so on. Not everyone believes the stereotypes, but most children and adults are aware of them. Regardless of whether we come to hold these stereotypes as strong convictions or merely as familiar-but-distrusted notions, knowledge of their content alone can bias our perceptions of stereotyped groups.
Such bias poses a serious problem for students who belong to stigmatized groups. Researchers have shown that stereotyped expectations shape social interactions and over time can result in the fulfillment of that stereotype. For example, if a student's social identity suggests high academic ability, interest, or potential, he or she may be treated accordingly by a teacher--receiving more attention, more challenging material, more patience--and over time, develop into the bright student the teacher initially imagined. By the same token, negative stereotypes can have the opposite effect, leading a teacher to create a less encouraging learning environment for students from stigmatized groups.
Recent research suggests that students need never encounter actual prejudice or differential treatment to be hurt by stereotypes. Just as mere knowledge of a stereotype can influence the thinking and behavior of teachers, it can also have a direct effect on students' perceptions of themselves. Students from stigmatized social groups are often bothered by the possibility that they will be viewed through the lens of a negative stereotype. As a result, they feel extra pressure not to fail, else their poor performance be perceived as evidence confirming the stereotype. This pressure may in turn impair their academic performance, thereby creating a self-fulfilling prophecy known as stereotype threat.
High school and college females tend not to perform as well as males on standardized math tests. Some have interpreted the performance gap as reflecting innate sex differences in mathematical ability. Others have argued that the gap is a consequence of childhood sex-role socialization patterns in which boys receive more encouragement than girls to participate in math activities. Neither of these accounts fares well in explaining why the performance gap emerges in high school. We contend that the gap is due in part to feelings of stereotype threat that female students may experience when they encounter difficulties in a math test. This contention is based on our observation that the largest gender gaps occur when females are tested in situations that make their gender identity especially salient.
For example, we have found that college women (but not men) taking a math test are affected by the gender composition of the testing room: When the majority of students in the test room are men, women tend to achieve lower scores than when the majority are women. Moreover, simply asking women to fill out a demographics question indicating their sex prior to taking the test (as all students do when they take the SAT, LSAT, GRE, etc.) can lower their test scores.
Given these findings, it's reasonable to ask whether instructors should make their female students aware of the stereotype threat phenomenon. This question raises another--does forewarning benefit students in this situation? It does in other situations. For example, forewarning can counteract persuasion, leading audiences to direct their thoughts inward to develop arguments countering those offered by a persuasive speaker. However, the principal symptom of stereotype threat is a pre-occupation with unwanted self-related thoughts (e.g., “girls can't do math”). People have a difficult time suppressing such thoughts. Sometimes, the attempt to suppress thoughts can have a rebound effect, leading people to think these thoughts even more than if they had not tried to suppress them.
We observed rebound of this sort in a simple experiment. Prior to administering a standardized math test (a quantatitive GRE subsection) to a group of undergraduate men and women college students, students read different messages about how to cope with test difficulties. One group was simply told to persevere when they encountered difficult problems on the test. A second group was told about the phenomenon of stereotype threat and instructed to suppress stereotype-related thoughts to reduce its effects. We observed the typical gender gap in performance (males achieved approximately 12% higher accuracy scores on average than females) in the control condition. The gap was even wider (about 20%) when females were forewarned about stereotype threat. Thus it appears that forewarning women about the phenomenon can actually aggravate the problem when they attempt to suppress relevant thoughts.
The news was not all bad, however. A third group of our participants was told about stereotype threat, but instead of being instructed to suppress thoughts, they were told of another stereotype that had different implications for their performance. Specifically, they were told that past research had demonstrated that students enrolled at elite private colleges (such as the students in our study) were far less vulnerable to stereotype threat than other students. Although this statement is groundless, it did prompt our female students to contemplate an aspect of their identity other than gender, one for which there are positive performance expectations. In this condition, the gender gap in performance was far smaller than the other two conditions (about 6%). In effect, by emphasizing another (but baseless) stereotype for females, the oft-observed gender gap in math test performance was reduced by a sizable margin.
Over the years, instructional communication researchers have identified a myriad of environmental factors that can render academic settings inhospitable to women, from overt sexist language to more subtle cues that trivialize women's intellectual performance. Consideration of these factors has informed the development of communicative strategies for raising educators' gender sensitivity in the classroom. However, the phenomenon of stereotype threat suggests that internal factors (e.g., one's fear of confirming a negative stereotype) can also contribute to female students' perception of an antagonistic academic environment. To address this problem, educators must not only avoid messages that devalue females' learning, but also encourage them to resist preoccupation with negative gender stereotypes. One promising strategy for achieving this goal is to remind female students of the many aspects of their identity that are associated with positive performance expectations--private school student, honor roll achiever, scholarship recipient, hard worker. Just as social identity appears to be the cause of stereotype threat effects, it can serve as the cure as well.