Lack of stereotype threat at a liberal arts college

College Student Journal, Sept, 2008 by Mark G. Rivardo, Michael E. Rhodes, Brandi Klein

Stereotype threat has been demonstrated to reduce the performance of stereotyped individuals in the threatened domain (Steele & Aronson, 1995). This study attempted to replicate the finding that stereotype threat instruction can erase the performance deficit women experience in math performance (Johns, Schmader, & Martens, 2005) and to further evaluate the arousal hypothesis of stereotype threat (e.g. Ben-Zeev, Fein & Inzlicht, 2005). The study provided no evidence of stereotype threat F(2, 91) = 1.60, p = .208, partial [h.sup.2] = .034. Stereotype threat may be less likely to affect performance at a small, liberal arts institution where the learning environment is both nurturing and personal.

**********

Stereotype threat occurs when members of a negatively stereotyped group are put in a situation where their performance on a given task could confirm the stereotype (Steele & Aronson, 1995). The pressure caused by this knowledge can hinder performance on the task and make confirmation of the stereotype more likely. For example, African-American participants performed worse when a task was described as being diagnostic of intellectual ability, than when it was simply described as an instrument for studying problem solving (Steele & Aronson, 1995). Researchers have also examined stereotype threat in other stereotyped groups and tasks, such as women and math (e.g. BenZeev, Fein, & Inzlicht, 2005; O'Brien & Crandall, 2003; Spencer, Steele & Quinn, 1999), and Black and White athletic performance (Stone, Lynch, Sjomeling, & Darley, 1999).

Research clearly indicates that the effects of stereotype threat can vary based upon individual differences in coping. Some participants engage in self-handicapping such as withdrawing effort (Steele & Aronson, 1995) or engaging in less practice under stereotype threat conditions (Stone, 2002) to provide an alternative explanation for poorer performance. Others may react to the situation with denial, and thus prevent the threat from affecting their performance (von Hippel, et al., 2005). Coping sense of humor can also guard against the negative effects of stereotype threat (Ford, Ferguson, Brooks, & Hagadone, 2004).

Studies conducted on the effects of stereotypes threat on women's math performance have identified additional parameters for the phenomenon to occur: GRE and SAT items are often used to measure performance because the male advantage in math performance is typically only found in problem solving tasks and does not appear until after the high school years (Hyde, Fennema, & Lamon, 1990) and stereotype threat decreases performance only if the task is sufficiently difficult to challenge the individual (O'Brien & Crandall, 2003; Spencer, Steele, & Quinn, 1999). For women completing math tests, the presence of a female role model who demonstrates strong mathematical ability can improve women's performance on a difficult mathematical test (Marx & Roman, 2002). Similarly, when female participants read about successful female role models, they did better on quantitative GRE sample items (McIntyre, Paulson, & Lord, 2003). In addition, when the experimenter told the experimental group of female participants that women made better participants in psychology experiments, they did better than females in the control group. Inzlicht and Ben-Zeev (2000) found women performed better when they were tested in same-sex groups of three than when they were tested with two men (experiment 1) and that when women were tested in same-sex groups of three, their performance did not differ from that of males (experiment 2). Aronson, Lustina, and Good (1999) suggest the stereotype threat effect is greatest for people who value the domain being tested. Subsequently, some researchers have limited their samples to students who met an SAT Math cutoff (e.g. Inzlicht, Aronson, Good, & McKay, 2006; Spencer et al., 1999), indicated they valued math ability on a questionnaire (e.g. Ben-Zeev, Fein, & Inzlicht, 2005; Inzlicht & BenZeev, 2003), or were likely to value math based upon their course enrollment (e.g. statistics students, Johns et al., 2005). Although research has demonstrated stereotype threat's effect on performance to be robust, there certainly are limits to its applicability.

Johns, Schmader, and Martens (2005) found the advantage male participants had over female participants on GRE word problems disappeared when the women were told about stereotype threat and that the anxiety they might experience during the test could be due to stereotype threat, rather than to any real difficulty with the items. Johns et al. claimed the performance of members in the threatened-instructed group was increased because they externalized the arousal. Their hypothesis is based upon the arousal theory of stereotype threat supported by others (e.g. O'Brien & Crandall, 2003; Blascovich, Spencer, Quinn, & Steele, 2001; Ben-Zeev, Fein, and Inzlicht, 2005), namely that stereotype threat activation leads to arousal that is attributed to difficulty with the test items. The decrease in performance caused by stereotype threat may be due to a reduction in working memory capacity as was found with women and Latinos under stereotype threat conditions (Schmader & Johns, 2003). This reduction in working memory capacity mediates the reduction in performance on standardized quantitative tests. Taken together, these ideas form a reasonable theory of stereotype threat. Stereotype threat activation leads to increased arousal that is attributed to difficulty with the test. The arousal leads to a reduction in working memory capacity, which makes the already difficult test even more difficult for the stereotyped individual, and therefore performance is hindered.