Sex stereotypes and implicit personality theory: Toward a cognitive—Social psychological conceptualization

Progress in understanding sex stereotypes has been impeded by the failure of researchers to address two critical conceptual questions: What is a sex stereotype? How do sex stereotypes function in social cognition and behavior? As a step toward answering the first question, the meaning of the term “sex stereotype” was considered. On the basis of points of agreement among extant conceptual definitions of the construct “stereotype” (in both the female-male and ethnic relations literatures), a generic definition of “sex stereotypes” is proposed: the structured sets of beliefs about the personal attributes of women and of men. In order to relate sex stereotypes more closely to research and theory on “normal” psychological processes, this basic definition is recast in terms of the person perception construct, “implicit personality theory”: the structured sets of inferential relations that link personal attributes to the social categories female and male. Two studies are presented to illustrate the utility of this translation. The remainder of the article addresses the second question. Here we offer preliminary ideas regarding a more general cognitive—social psychological framework for the study of sex stereotypes. Stereotype and stereotyping are distinguished, and each is discussed in light of relevant research in cognitive and social psychology.     

More than the Sum of Its Parts: Intersections of Sexual Orientation and Race as They Influence Perceptions of Group Similarity and Stereotype Content

Across two studies conducted in the United States, we apply an intersectional analysis to investigate how the interaction of sexual orientation (gay versus straight) and race (Black versus White) in male groups affect perceptions of inter-group similarity and stereotype content. The intersectionality hypothesis (Hypothesis 1) predicts that the interaction of race and sexual orientation will influence inter-group similarity perceptions and produce unique stereotype content that is not the result of simply adding race stereotypes to sexual orientation stereotypes. The related prototypicality hypothesis (Hypothesis 2) predicts that intersectional groups will generate more perceived dissimilarity and more unique attributes when they are non-prototypical with respect to race (Black) and sexual orientation (gay). We found strong support for intersectionality (Hypothesis 1); for Hypothesis 2, the prototypicality hypothesis, support was strongest with respect to the impact of non-prototypical sexual orientation (i.e., gayness). In Study 1, the group “Gay Black men” was viewed as most dissimilar to the group “Men,” and to its constituent ingroups. In Study 2, the group “Gay Black men” generated the most unique stereotype content. The addition of the label “gay” also reduced stereotype negativity, particularly in perceptions of Black men. The results contribute to research and theory by demonstrating that the intersection of ethnic and sexual orientation stereotypes is complex and produces meaningful differences in the perception of groups.

     

Sex stereotypes and implicit personality theory. I. A personality description approach to the assessment of sex stereotypes

It was proposed that sex stereotypes be phrased in terms of the person perception construct, “implicit personality theory,” as the structured sets of inferential relations that link personal attributes to the social categories female and male. The utility of this formulation was assessed by having 31 college students use a set of 66 personality traits to describe other people. A measure of trait co-occurrence was derived from these data and was used as input to Kruskal's multidimensional scaling program. A two-dimensional configuration was interpreted in terms of two orthogonal properties, Social Desirability and Potency. An Indirect Female-Male property (based on the proportion of times each trait was used to describe a male) was strongly related to the configuration and was closely aligned with the Potency vector. Thus, stereotypes of females and males were associated with the Potency dimension of person perception, with females seen as “soft” and males as “hard.” The directly rated sex property Male-Female was also located near the Potency vector, but was not strongly related to the configuration. This lack of fit may have been due to social desirability responding. The results support the utility of formulating sex stereotypes in terms of implicit personality theory and suggest the need to distinguish direct and indirect assessments of stereotypes.     

The conceptual structure of sex role stereotypes in college students

This study investigated sex stereotypes as natural language categories. Ten years ago, sex stereotypes were found to have a core-peripheral structure similar to that of the nonevaluative categories studies by Rosch (1973). Changes and stability in these conceptions were studied by having 56 female and 56 male undergraduate students rate 217 adjectives on Likert scales according to how well each adjective represented their own view of typical Men or Women. Mean ratings were analyzed to reveal a structure of core, peripheral, and nonmembers for Men and for Women. As was the case ten years ago, core traits reflected dimensions of “niceness-nurturance” for Women and “potency-strength” for Men. Female students also viewed Women as “competent,” while male students also viewed Women as “socially effective.” Adjectives that constituted the core for one category were generally placed in the periphery for the other category. As compared to ten years ago, the peripheries of the categories were expanded and showed a greater degree of overlap. College students' conceptions of Men and Women are best characterized as overlapping rather than as bipolar opposites. The formation of sex stereotype categories was discussed in terms of cognitive and social learning processes.     

Framing gender differences: Linguistic normativity affects perceptions of power and gender stereotypes

When unknown groups and equal status groups are compared by contrasting one group (“the effect to be explained”) against another (“the linguistic norm”), the group positioned as the norm is sometimes perceived as more powerful, more agentic, and as less communal. Such perceptions may contribute to status‐linked stereotypes, as group differences are spontaneously described by positioning higher‐status groups as the linguistic norm. Here, 103 participants considered gender differences in status to be larger and more legitimate and applied gender stereotypes more readily upon reading about gender differences in leadership that were framed around a male rather than a female linguistic norm. These effects did not generalize to 113 participants who read about gender differences in leisure time preferences framed around either norm. Jointly, these results suggest that the effects of linguistic framing on perceived group status and power and on group stereotypes generalize to domains where there are real differences in status, and contexts in which higher‐status groups are the default standard for comparison. Copyright © 2011 John Wiley & Sons, Ltd.     

Television and Sex-Role Stereotyping1

The possible influence of television on sex-stereotyped behavior was investigated in three studies. In Study I the portrayal of male and female central characters on children's Saturday morning television programs was examined, and a number of differences consistent with current sex-role stereotypes were found. Males and females were portrayed in different roles, they manifested different behaviors, and their behaviors were followed by different consequences. In addition, male characters were more frequent than females, and they exhibited higher rates of behavior. Similar differences in the portrayal of males and females in the commercial announcements accompanying these programs were found in Study II. The sexes differed in their frequency of appearance, their location, their roles, their expertise, and the consequences of their behavior. In Study III the effects on children's behavior of exposure to sex-stereotyped vs. non-stereotyped behavior by adult televised models were examined. It was found that children manifested greater imitation and recall for the behavior of a same-sex model with the result that boys exposed to “stereotyped” behavior by a male and female model manifested and recalled relatively more “masculine” behavior than those exposed to “non-stereotyped” behavior, while the opposite trend obtained for girls. Implications of these three studies for television's contribution to sex-stereotyped behavior are discussed.     

Stereotypes as Valid Categories of Knowledge and Human Perceptions of Group Differences

Stereotypes – beliefs about group differences – are more complex than is generally assumed. First, we address the multidimensionality of stereotypes under the framework of the Cubic EPA model which suggests that stereotypes are characterized by three dimensions: evaluation, potency, and accuracy. Specific attention is given to the relationship between stereotypes and totemic beliefs as collectively shared representations. Second, we review controversial research on the accuracy of stereotypes. Cultural stereotypes, personal stereotypes, judgment of individuals and groups, judgment criteria, and meta‐analytical results are examined, revealing the robust relationship between stereotypes and reality. Finally, we point to the importance of explanations of group differences, especially the perception of essence that is encouraged by group entitativity (perceptual “oneness”). We suggest that stereotyping is particularly powerful for groups with high entitativity and a perceived essence, and that a group's totem is the manifestation of the group's essence. Though we cannot resolve all the controversies relating to stereotyping, our perspective emphasizes stereotypes as categories useful for human interaction and survival.     

What's in a Name? The Effects of the Labels “Fat” Versus “Overweight” on Weight Bias1

This research examined the effects of the labels “fat” vs. “overweight” in the expression of weight bias, with the prediction that the label “fat” biases individuals to respond more negatively than does the label “overweight.” In Study 1, participants' attitudes toward people labeled as fat were less favorable than were their attitudes toward people labeled as overweight. In Studies 2 and 3, although participants chose similar‐sized figures to depict fat and overweight targets, weight stereotypes and weight attitudes were more negative toward people labeled as fat than those labeled as overweight. In addition, the endorsement of weight stereotypes mediated the biasing effect of the “fat” label on weight prejudice. Implications of this work for prejudice researchers and for public attitudes are discussed.     

A Confirmatory Factor Analytic Investigation of the TAS–20: Corroboration of a Five-Factor Model and Suggestions for Improvement

Two studies conducted around 1950 provided aggregate data on college students' attribution of characteristics to men or women (sex-role stereotype). Seeman (1950) provided data for three attributes clearly sex-role identified. Fernberger (1948) used a “story” with 16 blank spaces for response of “men” or “women” to assess sex-role stereotypes. Replications of these studies were conducted in undergraduate psychology classes during the 1970–71 school year to assess changes in sex-role stereotypes and in personal preferences. Results for stereotypes indicated either no change for the Seeman data or even more polarized stereotypes for the Fernberger data. No evidence for role convergence was found for stereotypes. Ss personal preferences were markedly discrepant from their stereotypic responses, as the literature would predict.     

Attraction increases and decreases as a function of emotion-attribution and appropriate social cues

To study the impact of startle-induced arousal on attraction, blindfolded subjects in a “vestibular function” study were startled by a loud noise accompanying the sudden backward tilt of the dental chair in which they were seated. In Study I, startled male subjects indicated (on a “postexperimental” questionnaire) greater attraction toward a pretty female experimenter than did control subjects. Study II demonstrated the reverse, with startled male subjects disliking a male experimenter more than controls. In Study III, female subjects startled by a male experimenter indicated greater attraction than controls, although the pattern of their responses differed from males. While an attempt to induce misattribution of arousal to a (placebo) pill (Study IV) or to a noise (Study V) with “arousal” side effects resulted in negligible attenuation effects on the startle—attraction relationship, in Study VI the imposition of a delay period between startle and experimenter ratings resulted in reduced ratings of attraction. The role of arousal in romantic attraction is briefly discussed, and the relevance of these data to theories of emotion is considered.     

Children's perceptions of relative competence in sex-typed occupations

Children ranging in age from 5 1/2 to 9 years old selected either a male or a female doll in response to questions about who would be “better” at an occupation. The occupations were selected so that half were female and half were male sex-typed occupations. The responses indicated that children have clear sex-typed expectations concerning occupational competence. While the differences between the “female” and “male” occupations were significant at each age level, the interaction revealed an increasing adherence to the stereotypes with increasing age. A sex of child main effect was produced by the tendency of female children to select the female doll more often than did male children. The results are interpreted as demonstrating that children's perceptions of these occupations reflect an evaluative bias and not simply a recognition of adult stereotypes or actual rates of employment in the positions.     

The definition of sex stereotypes via the adjective check list

The Adjective Check List (ACL) was employed in the empirical definition of male and female stereotypes by 50 male and 50 female college student subjects. Judgments by male and female subjects correlated highly. There were 33 male adjectives and 30 female adjectives on which at least 75% of both sexes agreed. With a 60% agreement criterion, there were 98 male and 83 female adjectives. Both male and female stereotypes were treated as hypothetical persons and were shown to be highly deviant on standard ACL norms, with the male stereotype being more deviant and perhaps more “disturbed” than the female stereotype. It was concluded that the ACL is a promising method for the definition and study of sex stereotypes.     

Are women invisible as leaders?

This study shows that women are unlikely to be seen as leaders. Subjects (n=448) rated each member of a five-person group (shown in a photograph) on leadership attributes and also chose one of the five as “contributing most to the group.” Eight different stimulus slides were used. In two slides the “head-of-the-table” cue to group leadership was pitted against sex-role stereotypes. A man seated at the head of the table in a mixed-sex group was clearly seen as leader of his group, but a woman occupying the same position was ignored. The head-of-the-table cue identified women as leaders only in all-female stimulus groups. The data were consistent with the hypotheses that sex stereotypes still control social judgments, and that discrimination operates nonconsciously and in spite of good intentions.     

Trait Attributions About College Students With a Physical Disability: Circumplex Analyses and Methodological Issues1

Trait attributions concerning able-bodied college students and those with a physical disability were investigated in two studies. In Study 1, 194 able-bodied students completed extensive adjective checklists in one of four experimental conditions: stimulus person physically disabled (wheelchair user) male, disabled female, able-bodied male, or able-bodied female college student. To avoid self-presentation biases, subjects completed checklists not in terms of their own views but in terms of commonly held stereotypes. Results showed that not only were fewer socially desirable and more undesirable traits attributed to students with a disability than to able-bodied students, but when tested for “sameness” vs. “oppositeness” using two circumplex models, traits attributed to students who have a disability were clearly the “opposite” of those attributed to able-bodied students. In Study 2, 115 students completed a trait checklist based on the findings of Study 1 with reference to one of the four stimulus persons. Although subjects reported their own views, the results were consistent with those of Study 1. It was also found that stereotyping in the socially desirable direction was related to stereotyping in the undesirable direction; both were related to lack of ease with students with a disability. Common stereotypes of wheelchair user students are listed and the implications of the findings for the design of programs to reduce prejudice and integrate students with a disability into academic life are discussed.     

The “making” of teenage pregnancy

I will do two things in this paper. First, I examine the issue of construction in the social sciences by using “teenage pregnancy” as an example. Following Michel Foucault's genealogical studies, I show that new categories were constructed to study teenage pregnancies, but that the construction involved does not support an extreme theory of construction—a theory which allows of nothing like “reality”. Second, I study the interaction between the categories used in investigations of teenage pregnancies and those to whom such categories are applied, namely the young mothers themselves. The interaction illustrates what Ian Hacking has described as a “looping effect” characteristic of categories used to study people and their behaviour. I claim that with teenage pregnancy, the looping effect is much more complex than Hacking suggests.     

Beautiful,Self‐Absorbed,and Shallow: People of Color Perceive White Women as an Ethnically Marked Category

Non‐Whites' stereotypes of White women were examined, comparing three perspectives: (1) White women are perceived similarly to ethnically “generic” stereotypes of women; (2) stereotypes are opposite of stereotypes of participants' own ethnic group; and (3) stereotypes are derived from media images of White women. In Study 1, participants listed stereotypes of White women in an open‐ended fashion. In Study 2, those stereotypes were developed into a close‐ended questionnaire, completed by a second set of participants. White women were perceived as attractive, blonde, ditsy, shallow, privileged, sexually available, and appearance focused. We concluded that White women are ethnically marked. Stereotypes of White women are consistent with media images of White women.