Current Research in Social Psychology

Normative fit (the consistency between stimulus persons’ features and perceivers stored stereotypes or group-based expectations) is known to facilitate social categorization. The present study expands upon earlier findings by showing that this effect extends to situations in which normative fit is based on generalized evaluation (i.e. evaluative fit). We presented 80 White participants with stimuli in which an equal number Black and White targets were disproportionately linked to stereotype irrelevant good and bad behaviors. In one condition, Black targets performed a disproportionate number of negative behaviors while White targets performed a disproportionate number of positive behaviors. In another condition, these proportions were reversed. Using the category confusion paradigm (Taylor, Fiske, Etcoff, and Ruderman, 1978), we found greater evidence of categorization use in the former condition than the latter condition. The use of this technique as an implicit measure is discussed. Current Research in Social Psychology (Vol. 12, No. 4) (Coats, Latu, & Haydel) 55 What determines whether a person is viewed as a group member or as a unique individual? While the consequences of social categorization have been widely studied (for example with respect to stereotyping, collective behavior, inter-group bias, [see Turner, 1987, for an overview]), the antecedents of social categorization have received comparatively less attention. Perhaps the most frequently cited factor said to facilitate social categorization is limited cognitive capacity (Coats and Smith, 2006; Gilbert and Hixon, 1991; Macrea, Milne and Bodenhausen, 1994). Since social categorization is often portrayed as providing a schema by which to organize and guide information processing, the need for such a mechanism is presumably most pressing when perceivers cannot devote the cognitive resources necessary to perceive social targets as unique individuals. In contrast to this resource-based explanation, self categorization theorists (Oakes and Turner, 1990) maintain that social categorization results not from limited cognitive resources, but rather from a concerted effort to make sense of the social world. They see social categorization as largely a function of category fit. Comparative fit refers to the extent to which the attributes and actions of the stimulus persons covary with their category membership. This type of fit has frequently been manipulated experimentally. For example, Oakes, Turner, and Haslam (1991, Experiment 1) manipulated comparative fit by presenting male and female stimulus person in such a way that male stimulus persons shared a specific attitude to a gender irrelevant issue while female stimulus persons shared a different attitude to this issue. In this study, categorization based on sex was more pronounced when fit was high than when attitudes were distributed randomly across categories. A second facet of category fit that is likewise thought to facilitate social categorization is normative fit. This is defined as the degree to which features of the stimulus persons are perceived to be consistent with stored stereotypes or normative expectations. Normative fit thus reflects the match between stimulus persons’ characteristics and the perceiver’s stereotype or expectations about the social category. For example, normative fit of the gender category would be strong when female targets discuss cooking and male targets discuss sports. Consistent with this position, a number of studies have found normative fit to increase activation and subsequent reliance on the corresponding category. For instance, Oakes, Turner and Haslam (1991 Experiment 2) presented six stimulus persons (three art and three science students) expressing either a stereotypical arts attitude (pro social life) or a stereotypical science attitude (pro hard work). In the normative fit condition, the art students advocated a pro social life position while the science students expressed a pro hard work attitude. In another condition, the attitude positions were reversed. The results of the study showed greater categorization based on sex in the former, normative fit, condition. In another study, Blanz and Aufderheide (1999) found the category "subject of study" to be more salient among participants presented with stimuli in which psychology students expressed pro alternative medicine statements and medical students expressed anti alternative medicine statements (the normative condition) compared to participants exposed to the opposite configuration. The paradigm that has been used most widely to assess social categorization is the category confusion task (otherwise known as the "who said what" task) developed by Taylor and colleagues (Taylor, Fiske, Etcoff, and Ruderman, 1978). This paradigm involves showing participants a stimulus video or computer presentation featuring behavioral descriptions or Current Research in Social Psychology (Vol. 12, No. 4) (Coats, Latu, & Haydel) 56 statements made by members from two different social categories (e.g., sex or race). Afterwards, participants are instructed to assign each behavior or statement to the person who had performed the behavior or expressed the statement. Incorrect responses are then coded as either withincategory errors (i.e., where the target-description match is wrong but the target chosen is a member of the same category as the correct target) or between-category errors (i.e. where the target-description match is wrong and the target chosen is from the other category). A higher proportion of within-category errors relative to between-category errors provides evidence that the social categorization was salient and used to organize information about the stimulus persons. Blanz and Aufderheide (1999) and Biernat and Vaescio (1993) used this measure of social categorization in their studies showing the facilitative effect of normative fit. We use this method in the present study. Previously published studies examining the effect of normative fit on social categorization have all done so by varying information about stimuli members with respect to specific stereotypic characteristics or, more typically, with respect to stereotypic positions on a particular issue (e.g., members of one category express pro social attitudes and members of the other category express pro hard work attitudes). In addition, these characteristics or attitudes have tended to be relatively neutral in value. In the present study, we broaden the manipulation of normative fit to reflect global evaluation. Rather than performing a behavior or expressing a position reflecting a stereotypic characteristic or attitude, the stimulus persons in the current study perform a wide spectrum of positive and negative behaviors that are stereotype irrelevant (i.e., not a part of the stereotype of either target category). The only semantic commonality among the behaviors is valence. By using only stereotype irrelevant behaviors that vary in valance, we are able to examine whether the effect of normative fit on social categorization extends beyond specific expectations to reflect more generalized beliefs about the target categories. Doing so requires only that the social categories we use differ with respect to perceivers’ perceptions about the tendency of members to engage in positive and negative behaviors. We refer to this special form of normative fit as evaluative fit. In addition to investigating whether evaluative fit facilitates social categorization use, our study suggests a new way to examine prejudice and stereotypes. Thus far, the primary focus of studies examining the relationship between normative fit and social categorization has been to demonstrate that normative fit is, in fact, an antecedent to social categorization (this is why studies have featured attributes that are known to be stereotypic of target categories). Once the stereotypic nature of the attributes are established (typically through pretesting), stimuli are configured to be either high or low in normative fit and then category salience is measured. We intend to reverse this analysis by suggesting that evidence of social categorization may be used to identify stereotypes or normative expectations people hold about given group. An illustration may be helpful. Imagine a study in which the stimuli are configured so that in one condition behaviors reflecting one characteristic (e.g., violent) are linked disproportionately to members one category (e.g., Muslims) and behaviors reflecting the opposite (e.g., peaceful) are linked disproportionably to members of another category (e.g., Christians). In a second condition the stimuli configuration is reversed. Higher rates of categorization in the former condition than the latter would suggest that perceivers view the former configuration as more normative than the latter (i.e., that Muslims are more violent than Christians). In this way, we may use evidence of social categorization to identify the stereotypes or the normative expectations perceivers hold Current Research in Social Psychology (Vol. 12, No. 4) (Coats, Latu, & Haydel) 57 about a group. The advantage of this technique over direct paper and pencil measures is that it allows one to assess prejudice and stereotypes that people may not readily admit to on explicit measures (Sigall and Page, 1971). In the current study, we limit our focus to that of generalized prejudice (as the primary goal of the study is to examine whether evaluative fit facilitates categorization use) but we note here that the same process could be used to examine particular stereotypes (as is illustrated in the above example).