Running head: THREATENING INTELLECTUAL ENVIRONMENT A threatening intellectual environment: Why females are susceptible to experiencing problem-solving deficits in the presence of males

Does placing females in environments in which they have contact with males, cause them to experience deficits in problem-solving performance? Is a situational cue, such as gender composition, sufficient for creating a threatening intellectual environment for females-an environment that elicits performance-impinging stereotypes? To explore these questions, participants completed a difficult math or verbal test in 3-person groups, which consisted of two people of the same sex as the participant (same-sex condition) or of the opposite sex (minority condition). Minority females experienced performance deficits in the math test only, whereas males performed equally well on the math test in both conditions. Further investigation showed that females’ deficits were proportional to the number of males in their surrounding. Even females who were placed in a mixed-sex majority condition (2 females and 1 male) experienced moderate but significant deficits. Findings are discussed in relation to theories of distinctiveness, stereotype threat, and tokenism. Threatening intellectual environment 3 A threatening intellectual environment: Why females are susceptible to experiencing problem-solving deficits in the presence of males Females currently comprise a small minority in the natural and physical sciences. A recent NSF report (1996) showed that in academia, females constitute 35% of undergraduate students enrolled in physics, math, and computer science classes, 16% of undergraduate students enrolled in engineering classes, and less than 10% of graduate students in physics and engineering. Moreover, this report shows that females suffer from higher attrition rates in their academic careers so that by the time women reach the workplace they occupy only 22% of jobs in mathematical and scientific domains. Does females’ problem-solving performance diminish when females are placed in an environment in which males outnumber them? If so, are such performance deficits specifically linked to domains, which are associated with negative stereotypes about females’ intellectual capacity? Examining these questions is informative both with respect to theories of how given social stereotypes affect the intellectual processing of individuals who are the targets of these stereotypes as well as to educational practice. Being outnumbered may cause females to suffer from stereotype threat, which is a situational phenomenon that occurs when targets of stereotypes alleging intellectual inferiority are reminded of the possibility of confirming these stereotypes (Aronson, Lustina, Good, Keough, Steele, & Brown, 1999; Aronson, Quinn, & Spencer, 1998; Spencer, Steele, & Quinn, 1999; Steele & Aronson, 1995; Steele, 1997). The experience of stereotype threat may, in turn, interfere with intellectual performance, especially when individuals are highly identified with success and achievement in the given domain. For example, Spencer et al. (Experiment 3, 1999) showed that high-achieving females performed significantly worse than males on a standardized math test when the stereotype about their math ability was made salient. Saliency was manipulated by informing participants that the test they were about to take had elicited sexdifferences in the past, presumably favoring males. These findings are relevant to explaining sex-differences on the math SATs (see Brown & Josephs, 1999), because females who are highly Threatening intellectual environment 4 identified with pursuing a college education may underperform as a result of a heightened test anxiety and loss of motivation, due to stereotype threat (e.g., Steele, & Aronson, 1995). The effects of stereotype threat on academic achievement have been demonstrated across different stereotyped groups, such as African-Americans (Steele & Aronson, 1995), Latinos (Aronson & Salinas, 1997), and students of low socioeconomic status (Croizet & Claire, 1998). There may be no need to elicit stereotype threat by “hitting participants over the head,” either by directly presenting people with the stereotype (Spencer et al., 1999), or by reminding participants of their membership in a stereotyped group (Steele & Aronson, Study 4, 1995). Instead, we contend that stereotype threat may be evoked by any factor that increases the saliency of group stereotypes. Deaux and Major’s interactive model of gender-related behavior (1987), suggests that the situation or environment can serve as a causal factor (in addition to attributes associated with the perceiver and the target), in determining whether gender stereotypes will be activated. We predict that one such environmental factor, being outnumbered by members of the opposite sex, will suffice to cause females to experience the detrimental effects of negative stereotypes about their mathematical ability. We refer to any such environment, which can activate the threatening effects of gender stereotypes as a threatening intellectual environment. The existence of this particular threatening intellectual environment for females is consistent with distinctiveness theory (Abrams, Thomas, & Hogg, 1990; McGuire, McGuire, & Fujioka, 1978; McGuire, McGuire, & Winton, 1979; McGuire & Padawer-Singer, 1976). This theory suggests that a minority status can evoke a sense of group identity, which is then incorporated into the working self-concept. For example, by manipulating the sex-composition of three-person groups, Cota and Dion (1986) found that whereas 34% of participants in the minority conditions noticed their own sex, only 16% of those in the same-sex condition did the same. Furthermore, McGuire, McGuire and Winton (1979) found that as the relative numbers of opposite-sex individuals increased, the spontaneous mention of gender increased proportionately. This last finding raises the possibility that as females are increasingly outnumbered by males, a situation which is common to many advanced-level quantitative high school classes, university Threatening intellectual environment 5 courses, and work place environments, females may become more aware of their gender. We predict that the increased awareness of one’s group, and the negative stereotypes that are associated with this group, will cause poorer performance. According to the theory of stereotype threat, a minority status should only disrupt cognitive functioning on a stereotyped task. However, according to tokenism theory (Lord & Saenz, 1985; Saenz, 1994; Saenz & Lord, 1989), a minority or token statusbeing a token minority in an otherwise homogeneous groupshould elicit cognitive deficits in all domains (Lord & Saenz, 1985; Saenz & Lord, 1989). Saenz (1994) argued that a token status invokes the feeling of being responsible for representing one’s minority group favorably in any given domain. Saenz suggested that the feeling of responsibility or self-consciousness diverts the token’s attention from the cognitive task at hand and therefore can result in deficits in problem solving and memory. Tokenism and stereotype threat make differential predictions regarding the effects of a minority status on cognitive performance (Steele & Aronson, 1995). Stereotype threat posits that it is the stereotype itself that causes evaluation apprehension. When people are placed in the minority and are asked to perform in a stereotyped domain, people are reminded of the stereotype that detracts from their group’s reputation. It is the cognizance of the stereotype that then leads to performance deficits. In contrast, tokenism posits that the token status itself causes selfconsciousness, regardless of whether individuals are asked to perform in stereotyped or nonstereotyped domains, and therefore feel the pressure of being responsible for representing other members of their group in a positive light. Thus, both stereotype threat and tokenism would predict that females would experience performance deficits when they comprise of a numerical minority. Stereotype threat, however, would make the qualification that a deficit would only occur in a negatively stereotyped domain, such as mathematics. Goals and the experimental paradigm. Threatening intellectual environment 6 The main goals of the current study were to test whether: (a) placing females in the minority, or in an environment in which they are outnumbered by males, is sufficient to create a threatening intellectual environment that causes deficits in females’ intellectual performance; and (b) whether minority-induced performance deficits, if any, are specific to a stereotyped domain or generalize to stereotyped and non-stereotyped domains just the same. The experimental paradigm consisted of asking participants to take a test with either two people of the same sex (same-sex condition), or with two people of the opposite sex (minority condition). The test comprised of items from either a stereotyped (math) or non-stereotyped domain (verbal). If being in the minority affects females’ performance negatively, then females in the minority condition would experience decrements relative to females in the same-sex condition, whereas males would perform similarly in both conditions. If performance decrements occur, and are mediated by stereotype threat, then they would only occur in the stereotyped domain (stereotype threat hypothesis). In contrast, if performance deficits occur as a result of being a token in a group, then they should occur in both stereotyped and non-stereotyped domains (tokenism hypothesis). Experiment 1 was designed to examine these hypotheses with female participants. Experiment 2 was crafted to control for the performance of male participants as well as to examine whether females’ math performance would decrease in proportion to the relative number of males who are present in their environment.