«A Resource Guide for Parents and Teens Developed and Compiled by the Youth Council of the DuPage Workforce Board A Letter to Parents: Your teen’s ...»
free of gender bias would remove the stereotype threat, and women would perform as well as men. If, however, gender differences in performance were due to sex-linked ability differences in math, women would perform worse than men even when the stereotype threat had been lifted. They found that women performed significantly worse than men in the threat situation and that the gender difference almost disappeared in the nonthreat condition (see figure 15).
In the ensuing decade more than 300 studies have been published that support this finding.
The results of these experiments show that stereotype threat is often the default situation in testing environments. The threat can be easily induced by asking students to indicate their gender before a test or simply having a larger ratio of men to women in a testing situation (Inzlicht & Ben-Zeev, 2000). Research consistently finds that stereotype threat adversely affects women’s math performance to a modest degree (Nguyen & Ryan, 2008) and may account for as much as 20 points on the math portion of the SAT (Walton & Spencer, 2009).
While 20 points on a test with a total possible score of 800 may seem small, in 2008 the AAUW 40 average male score on the SAT math exam was 30 points higher than the average female score, so eliminating stereotype threat could eliminate two-thirds of the gender gap on the SAT-M.
Aronson’s research also has shown that high-achieving and motivated women in the pipeline to STEM majors and careers are susceptible to stereotype threat. Aronson conducted a field experiment at a large public university in the southwest to investigate stereotype threat among students in a high-level calculus course that is a pipeline to future careers in science. The results showed no difference in performance between female and male STEM majors when they were told that a difficult math test was a diagnosis of their ability (threat condition);
however, when the threat was removed by telling the students that women and men performed equally well on the test, the women performed significantly better than the men (Good et al., 2008).
Stereotype threat also has implications beyond test performance. In an interview with AAUW, Aronson suggested that one reason girls lose confidence as they advance in school stems from “the stereotyping that students are exposed to in school, the media, and even at home that portrays boys as more innately gifted and math as a gift rather than a developed skill.
Without denying that biological factors may play a role in some math domains, psychology also plays a big role.” Additionally, a repeated or long-term threat can eventually undermine aspirations in the area of interest through a process called “disidentification.” Aronson describes disidentification as a defense to avoid the risk of being judged by a stereotype. Faced with a stereotype that girls are not good at math, for example, an individual might respond by claiming, “I don’t care about math; it’s not who I am.” In extreme cases, rather than repeatedly confronting a negative stereotype, girls and women might avoid the stereotype by avoiding math and science altogether.
Fortunately, Aronson and others have shown that stereotype threat can be alleviated by teaching students about it ( Johns et al., 2005), reassuring students that tests are fair (Good et al., 2003), and exposing students to female role models in math and science (McIntyre et al., 2003, 2005). Another promising approach draws on the work of Carol Dweck, profiled in the previous chapter. Encouraging students to think of their math abilities as expandable can lift stereotype threat and have a significant positive effect on students’ grades and test scores (Aronson et al., 2002; Good et al., 2003). In the interview with AAUW, Aronson stressed that “exposing students to role models who can help students see their struggles as a normal part of the learning process rather than as a signal of low ability” can boost the test scores of both minority students and girls.
• Teac h st udents and teac hers ab out stereot y p e threat.
Research with college students shows that acknowledging and explicitly teaching students about stereotype threat can result in better performance. Teachers and college faculty are best suited to do this and, therefore, need to be educated about stereotype threat.
Fewer girls than boys say they are interested in science or engineering careers (American Society for Quality, 2009; WGBH, 2009). The work of Shelley Correll, a sociologist at Stanford University, sheds light on how girls’ and women’s seemingly voluntary decisions to avoid STEM careers are influenced by the cultural belief that science and math are male domains.
Correll’s research focuses on self-assessment and its consequences for interest in math and science. She found that among students with equivalent past achievement in math, boys assessed their mathematical ability higher than girls did. Controlling for actual ability, the higher students assessed their mathematical ability, the greater the odds were that they would enroll in a high school calculus course and choose a college major in science, math, or engineering.
Correll found that boys were more likely than their equally accomplished female peers to enroll in calculus not because boys were better at math but because they believed that they were better at math. When mathematical self-assessment levels were controlled, the previous higher enrollment of boys in calculus disappeared and the gender gap in college major choice was reduced (Correll, 2001). In a follow-up study Correll (2004) verified in a laboratory experiment that when cultural beliefs about male superiority exist in any area, even a fictitious one, girls assess their abilities in that area lower, judge themselves by a higher standard, and express less of a desire to pursue a career in that area than boys do.
Undoubtedly, many factors influence an individual’s career choice, but at a minimum, individuals must believe they have the ability to succeed in a given career to develop preferences for that career. If girls do not believe they have the ability to become a scientist or engineer, they will choose to be something else. Correll’s research findings suggest that helping girls understand that girls and boys are equally capable in STEM areas will increase girls’ self-assessment of their math and science skills, which, in turn, will increase girls’ aspirations for careers in STEM fields.
Correll first became interested in the differences between boys’ and girls’ assessments of their science and math abilities when she taught high school chemistry for a few years before attending graduate school. She noticed that no matter how poorly the boys in her chemistry Shelley Correll is an associate professor of sociology at Stanford University. Her research examines how cultural 5 beliefs about gender influence educational and career paths. In addition to her work on self-assessment described in this chapter, her most recent project considers how stereotypical beliefs associated with motherhood influence the workplace evaluations, pay, and hiring of women who give evidence of being a mother.
AAUW 44 classes did, they continued to think that they were very good at chemistry; however, no matter how well the girls performed, it was difficult for Correll to convince them that they actually had some scientific ability. Once in graduate school Correll focused on how gender stereotypes attached to different skills or tasks influence how girls and boys understand their abilities independent of test scores or grades and how these gender differences in self-assessments contribute to gender differences in career choice.
S T E r E oT y P E S A n d S E l F - A S S E S S M E n T S
How do stereotypes affect self-assessments? Correll explains that we use stereotypes as “cognitive crutches” in situations in which we do not know how to judge our performance. Research shows that even individuals who do not personally endorse beliefs that men are better than women at math are likely to be aware that these beliefs exist in the culture and expect that others will treat them according to these beliefs. This expectation, or what we think “most people” believe, has been shown to influence judgments (Foschi, 1996; Steele, 1997; Lovaglia et al., 1998). If a girl believes that most people, especially those in her immediate environment, think boys are better than girls at math, that thought is going to affect her, even if she doesn’t believe it herself. Even if no one really believes that boys are better at math, the fact that a girl thinks they believe it is what matters. This is the reason that the 2005 comments of Larry Summers—the former Harvard president who famously doubted that women are capable of succeeding at the highest levels of science and engineering—were so damaging. Because he spoke from such a powerful position, his remarks gave credibility to the stereotype that women may lack the aptitude to succeed in STEM fields.
Correll published a study in 2001 that looked at the correlation between students’ math achievement and self-assessment of their math ability by gender and the influence that selfassessment has on persistence on a path to a STEM career. This study analyzed the National Educational Longitudinal Study of 1988 (NELS-88), a national dataset of more than 16,000 high school students. The first NELS-88 survey was conducted in 1988 when the students were in the eighth grade. A subsample of the original students was again surveyed in 1990, 1992, and 1994, when most were sophomores, seniors, and two years beyond high school, respectively.
Correll identified three items on the survey as indicators of mathematical self-assessment:
“Mathematics is one of my best subjects,” “I have always done well in math,” and “I get good marks in math.” Students were asked to agree or disagree, on a six-point scale, with these statements during their sophomore year of high school. Student mathematical achievement was approximated through past math test scores and average math grades that students received in high school. Correll’s analysis showed that high school boys were more likely Why So Few? 45 than their female counterparts of equal past mathematical performance to believe that they were competent at mathematics. Interestingly, the effect was reversed when the students assessed their verbal ability: female students made significantly higher self-assessments of verbal ability, controlling for actual verbal performance. This suggests that stereotypes about gender influence students’ perceptions of their abilities in particular fields: boys do not assess their task competence higher than girls do in every area, just in the areas considered to be masculine domains.
Most important for understanding how gender differences in self-assessment influence women’s underrepresentation in science and engineering, Correll’s research found that higher mathematical self-assessment among students of equal abilities increased students’ odds of enrolling in high school calculus and choosing a quantitative college major. In her sample, she found that boys were 1.2 times more likely than their equally capable female counterparts to enroll in calculus. Correll found this difference to be due to differences in self-assessment.
When girls and boys assessed themselves as equally mathematically competent, the gender difference disappeared, and girls and boys were equally likely to enroll in calculus. Likewise, 4 percent of female students compared with 12 percent of male students in Correll’s sample chose a college major in engineering, mathematics, or the physical sciences. Although controlling for mathematical self-assessment did not eliminate this gender difference in college major choice, it did reduce the difference. Together these findings suggest that cultural beliefs about the appropriateness of one career choice over another can influence self-assessment and partially account for the disproportionately high numbers of men in the quantitative professions, over and above measures of actual ability (Correll, 2001).
Interestingly, Correll found that young women who enrolled in high school calculus were about three times more likely than young women who did not take calculus to choose a quantitative major in college. In comparison, young men who enrolled in calculus were only about twice as likely as young men who did not take calculus to choose a quantitative major.
Thus it appears that taking calculus in high school is a better predictor of selecting a quantitative college major for women than it is for men. Another interesting finding was that higher verbal self-assessments decreased the odds of enrolling in calculus and choosing a quantitative major, indicating that students use relative understandings of their competencies when making career-relevant decisions. Lubinski and Benbow (2006) showed that girls who do very well at math are more likely than their male peers to do very well at verbal tasks as well. In addition to societal expectations, relatively strong verbal abilities may encourage mathematically talented girls to consider future education and careers in the humanities or social sciences rather than science and engineering fields.