Think women, think warm 1 Running head: Think women, think warm Think Women, Think Warm: Stereotype Content Activation in Women With a Salient Gender Identity, Using a Modified Stroop Task Judith B. White1 Wendi L. Gardner2 Address correspondence to Judith B. White Post 100 Tuck Hall, Hanover, NH 03755 Email [email protected] Fax 603.646.1308 Voice 603.646.9054 in press, Sex Roles 1 Tuck School of Business, Dartmouth College, Hanover, New Hampshire 2 Department of Psychology, Northwestern University, Evanston, Illinois Think women, think warm 2 ABSTRACT We examined whether a salient gender identity activates gender stereotypes along the dimensions of sociability and ability (Fiske, Cuddy, Glick, & Xu, 2002). A sample of US undergraduates (40 men, 38 women) instructed to think about women subsequently took longer to name the colors of words associated with sociability than ability on a modified Stroop task. Solo women in another sample of US undergraduates (45 women) showed the same response pattern. Women in a third sample of US adults (20 men, 16 women) showed a similar pattern. Meta-analysis of the three samples suggests women with a salient gender identity experience relative activation of only the positive dimension of a stereotype (e.g. "woman" equals warm). KEYWORDS: stereotype content, stereotype activation, gender, Stroop task, solo status Think women, think warm 3 Sugar and spice and all things nice, that's what little girls are made of. Mother Goose Introduction Professional success is a challenge for men and women alike. It takes hard work to earn respect and status regardless of whether one is male or female. Yet researchers have long theorized and even shown that women may have a slightly harder time achieving this success, because competent and powerful are attributes not consistent with the widely held stereotype of women (Eagly & Karau, 2002; Glick & Fiske, 1996). As in the nursery rhyme, women are considered warm and nice, not necessarily able and competent. An additional barrier for women is that as they advance in their careers, they are increasingly likely to be outnumbered by men in the executive suite (Helfat, Harris, & Wolfson, 2006). Being outnumbered heightens one’s self-consciousness with regard to minority status, and makes minority identity, e.g., being a woman, salient (Cota & Dion, 1986). Being in the minority places added pressure on women to conform to the female stereotype (Kanter, 1977), and raises the fear of backlash from male and female colleagues alike if they are not sufficiently "nice" (Rudman & Glick, 1999). We tend to think of these pressures as being directed toward a woman from her colleagues. However, stereotypes can have a negative effect on a woman's ability to advance in her career regardless of whether her colleagues are overtly sexist, whether or not they discriminate against women. This is because stereotypes can create an internal cognitive problem for women who aspire to high status positions in a male-dominated organization. Think women, think warm 4 Stereotypes are part of our cultural knowledge. The content of gender stereotypes is both explicit (people can tell us what they are) and implicit (people apply the information automatically when making judgments). Men and women alike learn the content of stereotypes and are affected by the information. When the female stereotype is cognitively activated, it affects how women view themselves, a phenomenon called self-stereotyping (Sinclair, Huntsinger, Skorinko, & Hardin, 2005). It may also cause women to behave in a manner more consistent with the stereotype, called an ideomotor (Wheeler & Petty, 2001) or active-self (DeMarree, Wheeler, & Petty, 2005) response. Research has shown that being in the minority results a women's gender stereotype being activated (Inzlicht, Aronson, Good, & McKay, 2006). Thus, when a woman works in a male-dominated environment, stereotype activation could affect her self-views and her behavior, and consequently her performance, even when her colleagues are not overtly sexist or discriminatory. The purpose of the present research was to take the first look at how the positive (sociability) and negative (ability) dimensions of the female gender stereotype as described in stereotype content model (Fiske, Cuddy, Glick, & Xu, 2002) are activated in women who work in male-dominated environments. We used a modified version of a Stroop test (1935) to measure stereotype activation. Words associated with the underlying dimensions of gender stereotypes (sociability, ability) were presented along with non-trait words in colored fonts on a computer screen. Participants were instructed to ignore the word meaning, and to respond to the font color as quickly as possible by pressing a color-coded key. In this implicit measure of stereotype activation, a measurable delay in response times is an indication that stereotype words are associated with an activated construct which is interfering with the assigned task of color identification. Three studies used the modified Stroop test to examine the pattern of gender Think women, think warm 5 stereotype activation. Gender category salience was manipulated in Study 1 by explicitly instructing participants to think about men or women. Gender identity salience was manipulated in Study 2 by giving some women solo gender status. Gender identity salience was then assumed in Study 3 by testing a sample of adult women in the workplace. This paper extends previous work in two ways. First, we examine both the positive and negative content of gender stereotypes (e.g., women are good at nurturing but bad at math) and test whether one or both dimensions are activated when a woman's gender identity is salient. Second, we use a modified Stroop task (Stroop, 1935) to measure cognitive interference in contrast to previous research that has used implicit measures such as the IAT (Greenwald, McGhee, & Schwartz, 1998) or word-stem completion tasks (e.g., Steele & Aronson, 1995) that measure facilitation effects of stereotype activation. The results of these studies are important because they suggest that women who work in male-dominated environments may subconsciously think and possibly automatically conform to the "women-are-warm" positive stereotype dimension even when they do not face explicit stereotype threat. Identity Salience and Stereotype Activation Social identities are parts of the self-concept that arise from the knowledge that we share membership in a social group or category (Tajfel, 1981). Woman and man are social identities. It is generally accepted that the social context can make a particular social identity salient (Shih, Pittinsky, & Ambady, 1999). For example, being the only woman in a group makes that distinctive gender identity salient (Cota & Dion, 1986). Women who are in the numeric minority in their work environment report that they are strongly conscious of their gender identity and how it differentiates them from their co-workers (MacCorquodale & Jensen, 1993; Theberge, Think women, think warm 6 1993). A salient identity has a greater influence on our cognition and behavior, on "who we are" at any given moment (Gardner, Gabriel, & Lee, 1999). One way to think of a social identity is as a cognitive schema that contains identity- relevant information (Greenwald, 1980). Gender schemas contain gender-relevant information, including the content of gender stereotypes. Therefore, when a gender identity is salient, the associated gender stereotype is activated. An activated stereotype causes self-judgments, attitudes, and behavior to become more consistent with the stereotype. In the case of self- judgments, Sinclair, Hardin & Lowery (2006) show that Asian-American women describe themselves as better at math when their ethnic identity is salient, and better at verbal ability when their gender identity is salient. They also find that European-American men describe themselves as being better at math when their gender identity is salient, and better at verbal ability when their ethnic identity is salient. Female students who expect to meet with men at a later date describe themselves as more feminine than when they expect to meet with other women (Chiu et al., 1998). A salient gender identity also makes women's attitudes more stereotypical: more positive toward arts and more negative toward math (Steele & Ambady, 2006). Finally, an activated self-stereotype also affects behavior. Shih, Pittinsky and Ambady (1999) find that Asian-American women perform better on a math test when their Asian-American identity is salient, and worse when their gender identity is salient. The nature and direction of effects of stereotype activation in these studies depend explicitly on which social identity is salient (e.g., Asian or female), but implicitly on which dimension of the relevant stereotype is activated (e.g., good at math or socially reserved, versus bad at math or sociable). Understanding which dimension of the stereotype is activated is thus key to predicting the impact stereotype activation will have on an individual. Think women, think warm 7 Research has demonstrated that the same social identity can have opposite effects on behavior, depending on what part of the stereotype is activated. Consider that the stereotype of old people contains positive information about wisdom but also negative information about forgetfulness. When the positive dimension is primed elderly people show memory improvements, but when the negative dimension is primed they show memory deficits (Levy, 1996). Stone, Lynch, Sjomeling, and Darley (1999) find that white men perform well on an athletic assessment when it is presented as a test of intelligence, but poorly when the same assessment is presented as a test of athletic ability. In these studies, researchers control whether a positive (e.g., wise, intelligent) or negative (e.g. forgetful, un-athletic) dimension of a stereotype is activated, and consequently control how the stereotype activation affects performance. But we know little about what particular aspects or dimensions of gender stereotypes are activated when a woman's gender identity is salient in a work context.