تفاوت های فردی در قبول واقعیت تصورات قالبی

Abstract Previous research has documented individual differences in a range of constructs relating to social stereotyping, prejudice, and intergroup attitudes. However, research has not sought specifically to measure a general acceptance of social stereotyping. In the present research, we explored attitudinal, cognitive, emotional, and personality correlates of a person’s self-reported willingness to rely on stereotypical information when interacting with people of different social and cultural groups. In six studies (N = 1080) we found that more acceptance of stereotyping was associated with more explicit and implicit stereotyping of particular groups, less liberal gender-role values, more authoritarian attitudes, preference for hierarchies, higher social dominance orientation, less universal outlook, less complexity in describing others’ emotions, less utilization of emotional information, and more utilization of social categories (gender and race) when rating the similarity of faces, less agreeable and more agentic personality, and more rigid and simplistic cognitive style (all independent of one’s gender). Female and African-American participants were less accepting of stereotyping than male and Caucasian participants. The general tendency to accept stereotyping in daily life is a measurable individual difference that may prove useful in social-personality research.

1. Introduction Many people would agree that because stereotypes about social groups may be inaccurate and negative, relying on them in daily life may be inappropriate. However, even the most liberal-minded people engage in stereotypical thinking, probably far more often than they would like to admit. In daily life, we often feel that knowing something about another person’s social and cultural group memberships helps us to interact with that person, and the lack of such knowledge can produce uncertainty about how to behave and may undermine our feeling that we know the other person. Furthermore, uncertainty about others’ group memberships (whether these refer to gender, sexual orientation, ethnicity, social class, religion, political party, social clique, etc.) can make a person anxious about making social gaffes. Stereotype application may also seem to have a certain functionality to it, insofar as a simplifying assumption is at least a starting point from which to plan behavior toward another person. From the perspective of the person who has to make behavioral choices, using stereotypes—that is, making assumptions about members of social and cultural groups—may seem like a useful, sensible, and adaptive thing to do, and such a person might, as a consequence, feel a strong need to know about others’ social and cultural memberships as a guide for his or her own behavior. On the other hand, another person may be highly doubtful of the utility or validity of stereotypes and be more willing to build knowledge of others from the ground up. Thus, any attempt to assess the extent to which stereotyping occurs must consider the issue of individual differences in stereotyping. In the present research, we were interested in individual variation in the tendency to accept stereotyping in daily life. Acceptance of stereotyping as a general tendency not aimed at any particular group is an individual difference that could have considerable relevance for research on personality and social behavior. Acceptance of stereotyping as a general trait has been central to much theorizing (e.g., Allport, 1954), and there are numerous measures of stereotype application and prejudice (e.g., Brigham, 1993 and Glick and Fiske, 1996) or willingness to show prejudices toward specific groups (Dunton and Fazio, 1997 and Plant and Devine, 1998). Research also exists on personality types (e.g., Altemeyer, 1981) and other individual difference variables associated with stereotyping (e.g., Moskowitz, 1993). However, no research that we have found has examined the extent to which individuals explicitly report their general willingness to use stereotypical information in the course of daily life. Evidence does exist, however, to suggest that this might be a general trait among adults, as indicated by positive correlations among prejudices towards several different social groups (Bierly, 1985). Stereotyping has been defined in numerous ways throughout the history of research conducted on the topic (Lee, Jussim, & McCauley, 1995). Thus, it is important to be clear that in the current investigation we define acceptance of stereotyping as the belief that social and cultural group differences exist, comfort with thinking about groups in abstract terms, willingness to use information about group memberships in conducting interpersonal relations, and the belief that stereotypes are useful, essential, and relatively harmless in daily life. 1.1. Overview of present research and predictions Because no measure existed that captures the general tendency to accept stereotyping, we developed one and administered it in six studies of college students along with other measures that fell into five general categories: stereotyping and prejudice, social/interpersonal attitudes, cognitive style, emotion-related measures, and personality. In addition, the participants’ gender and ethnicity were measured. Following is a brief description of each of these categories of measures and the predicted relationships with acceptance of stereotyping. Members of social/cultural groups who have been the target of negative stereotypes and prejudice have been found to be less likely to express prejudice or discrimination (Altemeyer, 1998 and Whitley, 1999). Similarly, women have been shown to have less negative or stereotypic attitudes toward women than men do (Bierly, 1985, Glick and Fiske, 1996 and Swim et al., 1995) and less prejudice than men towards homosexuals, African Americans, old people, Jews, and Catholics (Allport and Kramer, 1946, Bierly, 1985 and Carter, 1948). We predicted that African Americans and women would score relatively low on acceptance of stereotyping. General willingness to endorse use of stereotypical information in making judgments about others implies the tendency to think stereotypically about specific groups, specifically African Americans and women in the present research. Regarding racial attitudes, we predicted that more acceptance of stereotyping would be associated with more negative attitudes measured both explicitly (measured with paper-and-pencil scales) and implicitly (measured with reaction times). Regarding sexist attitudes, we predicted that more acceptance of stereotyping would be associated with both negative and positive (idealizing) forms of stereotyping, and also less liberal (i.e., favorable to women) gender-role values. Acceptance of stereotyping was also predicted to be related to social/interpersonal attitudes. We predicted that more acceptance of stereotyping would occur in people who believe there should be hierarchies between social/cultural groups and among individuals, and who endorse more authoritarian values, because such beliefs and values reflect the conviction that group differences are real and important and also because stereotypes can be used in the service of reinforcing social hierarchies (Moskowitz, 1993). We predicted that people scoring higher on acceptance of stereotyping would have less trust in others and profess less faith in the (positive) universalism of human nature, consistent with the possibility that holding many stereotypes may reflect a generally misanthropic outlook. People scoring high on acceptance of stereotyping were also expected to see personality as fixed rather than malleable, consistent with the reasoning of Levy, Stroessner, and Dweck (1998) regarding the relation of entity thinking to the endorsement of specific group stereotypes. Levy et al. proposed that an entity approach to personality reflects the belief that individual traits are fixed, useful, and predictive, just as the acceptance of stereotyping suggests similar beliefs about groups. Finally, we predicted that people highly accepting of stereotyping would view their own outcomes as more controlled by powerful other people, consistent with both the hierarchical and somewhat suspicious outlook hypothesized earlier. Variables relating to cognitive style were also examined. People differ in how much they prefer quick, simplistic ways of thought and rigid, structured rules for living. Stereotypes have traditionally been defined as simplistic overgeneralizations (Lee et al., 1995), and there is indeed evidence that a person with a rigid or simplistic way of thinking would also be prone to use stereotypes (Schaller et al., 1995 and Webster and Kruglanski, 1994). Therefore, it was expected that acceptance of stereotyping would have a positive relationship with a cognitive processing style that is more simplistic (i.e., less elaborative) or categorical in nature. Another cognitive style variable we included was the need to evaluate, an individual difference variable that Jarvis and Petty (1996) predicted might be related to stereotyping. Another category of predicted correlates relates to attending to the emotional qualities of people and objects. Because paying attention to others’ emotions requires one to attend to transient states in others, such attention runs logically counter to the notion that others are homogeneous members of their category. Thus, attending to individuals’ emotions can be seen as intrinsically individuating and antithetical to the notion of stereotyping. Consistent with this, it has been found that individuating outgroup members reduces outgroup bias (e.g., Bettencourt, Miller, & Hume, 1999). It has also been suggested that one way to denigrate outgroups may be to deny that they experience subtle emotions (Leyens et al., 2000). We therefore speculated that there may be an inverse relation between acceptance of stereotyping and several variables suggestive of an interest in others’ emotions. We administered a behavioral (not self-report) measure of how much a person used emotion (versus race and gender) as a grouping dimension when rating the similarity of pairs of faces. We also administered questionnaire-based instruments that assessed how complexly a person described others’ imagined feelings and how much a person reported being able to get others to “open up” in conversation (where “opening up” implies emotional disclosure). We also measured how much participants projected emotion into interpretations of abstract drawings. If the proposed reluctance of people high in acceptance of stereotyping to deal with emotions is a general trait, we would predict a negative relation with this variable, but if their reluctance pertains to people as targets (not abstract line drawings), then we would predict no relation. Finally, to gain further insight into the personalities of people low versus high on acceptance of stereotyping, the Big-Five traits of extraversion, agreeableness, neuroticism, conscientiousness, and openness to experience were measured, as well as self-esteem, agency, and communion. No specific predictions were made regarding these personality variables. However, among men, traditional masculinity was associated with more negative attitudes about race and gender equality in the research of Wade, 2001 and Wade and Brittan-Powell, 2001, suggesting that an association might emerge between agency and acceptance of stereotyping. Because the basic methodology and some of the instruments were the same across studies, we describe all methodology in one section rather than present each study separately, and we group the studies’ results together thematically. When the same measures were used in more than one study, we report meta-analytic summaries (Rosenthal, 1991).

3. Results 3.1. Gender differences Men were significantly more accepting of stereotyping than women in all six studies (Table 3). In the table, the difference is expressed as the point-biserial correlation between gender and acceptance of stereotyping, which shows the magnitude of effect along with the same p value that a t test between males and females would have (the male and female means are shown in Table 2). Across the six studies, a meta-analytic summary ( Rosenthal, 1991) found an unweighted mean correlation of .26, a weighted (by sample size) mean correlation of .24, and a combined Z (Stouffer method; Rosenthal, 1991) of 7.72, p < .001. Thus, the gender difference was of moderate magnitude, very consistent, and highly significant statistically. All subsequent correlations shown in Table 3 controlled for gender using partial correlation. Table 3. Correlates of acceptance of stereotyping Category and measure Study Pearson correlation N I. Sociodemographic variables Gender (male = 0, female = 1) 1 −.25*** 205 2 −.21*** 229 3 −.24*** 412 4 −.26* 72 5 −.32 ** 62 6 −.31* 67 Ethnicity (African American, 0; Caucasian, 1) 3 .14* 255 II. Gender-related measures Benevolent sexism 1 .21*** 201 2 .17** 228 3 .19*** 360 Hostile sexism 3 .30*** 360 Modern sexism 3 .18*** 360 Traditional gender-role values 3 .27*** 360 III. Race-related measures Modern racism 3 .24*** 360 Attitudes towards blacks 6 .51*** 67 Implicit association task 6 .21+ 67 IV. General social/interpersonal attitudes Authoritarianism 3 .15** 360 Social dominance orientation 1 .41*** 201 2 .52*** 228 Preference for hierarchies 4 .45*** 72 Lack of universalism 1 .31*** 201 2 .22*** 228 Lack of trust in others 4 .33** 72 Fixed view of human nature 3 .11* 360 Control by powerful others 4 .35** 72 V. Cognitive style Low need for cognition 1 .14* 201 2 .10 228 Need to evaluate 3 .13** 360 Need for structure 1 .08 201 2 −.04 228 4 .28* 72 VI. Emotion-related measures Use of emotion in rating similarity of facesa 5 −.27* 62 Complexity of describing others’ feelings 5 −.43* 26 Sees emotion in abstract drawings 5 −.13 37 Gets others to “open up” 5 −.25 36 VII. Personality Self-esteem 4 −.01 72 Agreeableness 5 −.43** 33 Conscientiousness 5 −.22 33 Extraversion 5 .30+ 33 Neuroticism 5 .29 33 Openness 5 −.04 33 Agency 5 .56*** 38 Communion 5 −.22 38 Note. All correlations are partial correlations that control for gender. a Correlation for use of gender in rating similarity of faces = .26, p < .05; correlation for use of race in rating similarity of faces = .26, p < .05. * p < .05. ** p < .01. *** p < .001. + p < .10. Table options 3.2. Ethnic group differences In Study 3 (the only study large enough for such a comparison), acceptance of stereotyping was compared between African American (n = 20), Asian American (n = 22), and Caucasian (n = 235) participants (other ethnic groups were too small to include). The overall effect for ethnicity was F (2, 274) = 2.85, p = .059 (M African American = 1.63, M Asian American = 2.03, and M Caucasian = 2.03). Focused comparisons showed that African Americans were less accepting of stereotyping than Caucasians, t (253) = 2.38, p < .05. 3.3. Gender-related attitudes As shown in Table 3, people who were more accepting of stereotyping had significantly more benevolent sexist attitudes toward women (Studies 1–3), with a meta-analytic summary showing an unweighted mean correlation of .19, a weighted (by sample size) mean correlation of .19, and a combined Z of 5.28, p < .001. Also, people higher on acceptance of stereotyping scored significantly higher on hostile sexist attitudes toward women, scored higher on the Modern Sexism Scale, and had more traditional (i.e., less favorable to women) gender-role values (all Study 3). 3.4. Racial attitudes Table 3 shows that people who were more accepting of stereotyping scored higher on the Modern Racism Scale and held more explicitly anti-black attitudes as assessed by the Attitudes towards Blacks Scale. In addition, the implicit Race-IAT was marginally significantly related, such that people holding more negative implicit associations toward Black people tended to be more accepting of stereotyping. Most likely, there was not a more significant relationship between the two measures due to the Acceptance of Stereotyping Questionnaire assessing an explicit attitude reflecting an intentional preference to use stereotypes while the Race-IAT assesses an implicit tendency to make certain types of associations. Since these measures are assessing different constructs, it is understandable that the correlation between them would not be greater. The explicit and implicit attitudes instruments were not significantly related to each other, controlling for gender, partial r(65) = .14, p < .27. 3.5. Social–interpersonal attitudes Table 3 shows that people who were more accepting of stereotyping were significantly more likely to believe that certain groups deserve to be dominated (social dominance orientation, Studies 1–2), with a meta-analytic summary showing an unweighted mean correlation of .47, a weighted (by sample size) mean correlation of .47, and a combined Z of 9.60, p < .001. They were significantly more likely to believe that people are not all alike (universalism, Studies 1–2), with a meta-analytic summary showing an unweighted mean correlation of .26, a weighted (by sample size) mean correlation of .26, and a combined Z of 5.43, p < .001. They also scored significantly higher on authoritarianism, the fixed (entity) view of human nature, and preference for hierarchies; had significantly lower trust in others; and felt more controlled by powerful others. 3.6. Cognitive style As shown in Table 3, need for cognition had an inconsistent relation to acceptance of stereotyping in terms of significance tests, but when the p-values of Study 1 and 2 were combined meta-analytically (Stouffer method; Rosenthal, 1991), the overall relation was significant in showing the greater acceptance of stereotyping was associated with lower need for cognition, combined Z = 2.45, p < .01, one-tail, unweighted mean correlation = .12, weighted (by sample size) mean correlation = .12. Personal need for structure also had an inconsistent relation, but when combined meta-analytically across Studies 1, 2, and 4, it was related to acceptance of stereotyping such that greater acceptance was associated with a greater personal need for structure, combined Z = 1.68, p < .05, one-tail, unweighted mean correlation = .11, weighted (by sample size) mean correlation = .05. Higher need to evaluate was also significantly related to greater acceptance of stereotyping (Study 3). 3.7. Emotion-related measures The Similarity Rating Task (Study 5) measured participants’ preference for using emotion, gender, or race when rating the similarity of faces. Table 3 shows that greater use of emotion was significantly negatively related to acceptance of stereotyping and, correspondingly, greater use of gender and race was positively related to acceptance of stereotyping. Because of the ipsative nature of the task (i.e., using one dimension more required using one or both of the other dimensions less), scores on the similarity rating task were highly related to each other: participants’ use of emotion in the similarity rating task was negatively related to use of gender, r (60) = −.96, p < .001, and to use of race, r (60) = −.98, p < .001, and use of gender and race were positively related to each other, r (60) = .87, p < .001. Thus, use of the race and gender categories were used to the exclusion of using emotion, and vice versa. The results show that the tendency to use information about social categories (race and gender) and the tendency to use emotion had opposite relations to acceptance of stereotyping. People scoring higher on acceptance of stereotyping were significantly less likely to describe the imagined emotions of others in a complex way than people scoring lower (Table 3). The self-rated ability to get others to “open up” through self-disclosure was also moderately negatively related, though not significantly so. The tendency to see emotions in abstract drawings was not related to acceptance of stereotyping (all from Study 5). 3.8. Other personality variables Self-esteem was not related to acceptance of stereotyping (Table 3, Study 4). Results from Study 5 showed that less agreeable personality, more agentic personality, and to some extent more extraverted and neurotic personality were associated with more acceptance of stereotyping; agency was significantly related not only when gender was controlled for, but also for men and women separately—for men, r (14) = .59, p < .05, and for women, r (20) = .52, p = .01. 3.9. Socially desirable responding In Study 3, the correlation between acceptance of stereotyping and the Marlowe–Crowne Social Desirability Scale was r (421) = −.28, p < .001, suggesting that people who endorsed stereotyping were less interested in looking good than people who did not (conversely, that people who disavowed stereotyping were more interested in looking good). Importantly, when social desirability was partialed out of the Study 3 correlations shown in Table 3 (along with gender), the results were indistinguishable from those shown in the table. Thus, social desirability did not bias relations between acceptance of stereotyping and other variables in Study 3. Social desirability was examined in Study 5 as well. The correlations for the two subscales of the Balanced Inventory of Desirable Responding were as follows: Impression Management, r (55) = −.33, p < .05, and Self-Deceptive Enhancement, r (55) = −.17, ns. These correlations suggest that people who score high on acceptance of stereotyping are less interested in manipulating others’ impressions in a favorable way than people who score low, who may be motivated to present a socially desirable image of themselves, but those scoring high on acceptance of stereotyping do not differ appreciably in the extent to which they have a falsely positive self-view.