The role of women's identification with math and academic major in women's susceptibility to stereotype threat and stereotype lift

Abstract

A stereotype threat (ST) occurs when individuals underperform in a domain, for example in math, as a result of exposure to a relevant negative stereotype. Women engaged in math-intensive tasks can experience ST when negative stereotypes about women's math ability are made salient, via for example, test instructions that allege superior math performance by men. Evidence regarding the role of ST test instructions on women's math performances has been mixed (e.g., Bell, Spencer, Iserman, & Logel, 2003; Schmader, 2002). While prior studies found that women underperform in ST conditions that emphasize the validity of a math test (i.e., when a math test is presented as indicative of math ability), no study has included a condition in which the validity of a math test is downplayed (i.e., "test not indicative of math ability" conditions). Studies examining conditions alleging men's superiority in math (i.e., "men perform better than women" conditions) have not included conditions that presented a math test as indicative of math ability (Cadinu, Maass, Frigerio, Impagliazzo, & Latinotti, 2003; Johnson, Bernard-Brak, Saxon, & Johnson, 2012). Additionally, it is unclear which women are most vulnerable to math ST conditions. While ST is found to have greater impact on women who are highly identified with math relative to women with low identification with math (e.g., Aronson, Quinn, & Spencer, 1999), there is also evidence that women in math-intensive majors (e.g., engineering) have lower susceptibility to math ST than women not in math-intensive majors (e.g., psychology) (Crisp, Bache, & Maitner, 2009; Croizet et al., 2004). Furthermore, the roles of identification with math and academic major have been researched independently. The present study examines the roles of women's identification with mathematics and college majors on their susceptibility to math underperformance under two ST conditions, one related to the validity of the math test and the other involving comparisons in math performance between women and men. Women (n = 847), of whom 231 were in math-intensive majors and 616 were not in math-intensive majors at a large Mountain West state university, completed the Identification with Math Scale and reported their college majors five to seven days before completing a mathematics test. They were then randomly assigned to one of six math ST conditions in a 2 (Validity of Math Test Variable: test indicative of math ability, test not indicative of math ability) × 3 (Women-Men Math Performance Differences Variable: men perform better than women, no mention of differences in math performance, or women perform better than men) factorial design experiment. It was hypothesized that women in the "men perform better than women" condition would underperform relative to women in the "no mention of differences in math performance" condition. It was also hypothesized that women high in identification with math who were assigned to the "test indicative of math ability" condition would experience greater math underperformance than women in the "test not indicative of math ability" condition. A significant interaction between the Women-Men Math Performance Differences Variable and the Identification with Math Variable was found. Women high in identification with math in the "men perform better than women" condition scored significantly lower than women in the "no mention of differences in math performance" condition. No such difference in performance was observed for women low in identification with math. Women in the "women perform better than men" condition performed better than women in other conditions regardless of their identification with math. This study's findings suggest that women who strongly identify with math may be especially vulnerable to ST, consistent with past findings (e.g., Steinberg, Okun, & Aiken, 2012). In support of findings from past studies (e.g., Johnson et al., 2012), this study also demonstrates that to do well in math tests women may benefit from exposure to information explicitly contradicting female math incompetence stereotypes. Current study's findings have implications for intervention programs with highly math-identified women.