Category-Based and Person-Based Processing

Stereotyping represents the application of a social category (e.g., African Americans, Democrats, Jews, women) to someone rather than considering the unique aspects of the individual’s personality, attitudes, beliefs or behavior. A reliance on stereotypes, and adoption of a simplified view of others, may result in perceptual or inferential errors that undermine understanding one’s social environment. A more differentiated view may be achieved through the application of specific subcategories, or subgroups, but even here the person is viewed as a member of a general category rather than as a distinct individual.

It is reasonable to postulate that as we develop a more differentiated view of a person, and come to participate in a greater number and variety of shared experiences, stereotypes would become less useful and less readily applied in new thought or interactions. A more articulated concept would provide advantages due to its better fit to observed behavior and its greater inductive or predictive value. This better understanding could then foster more favorable interactions and might, under certain conditions, even facilitate better intergroup relations through reductions in prejudicial beliefs and attitudes (for example, see Pettigrew & Tropp, 2000 for discussion of the contact hypothesis).

This process of acquaintanceship and interpersonal learning has been viewed somewhat differently by different authors. Susan Fiske and her colleagues (Fiske & Neuberg, 1990; Fiske, Lin & Neuberg, 1999) proposed a single continuous process in which a person is first viewed in categorical terms, using a stereotype or social category, but alternatively with a subcategory, exemplar, or in a more piecemeal fashion, paying attention to the individual’s unique characteristics. The perceiver starts with the most general case of the general stereotype and proceeds to more specific representations until a good fit is achieved. If no category is effective in handling incoming information then piecemeal, attribute-based encoding will result. Focusing on the specific attributes of the person, and rejecting categorization based on broad social categories is referred to as individuation. In addition to goodness of fit, the adopted approach will depend on various motivational factors, such as self-relevance (Fisk & Neuberg, 1990), one’s goals (Stangor et al., 1992), contextual demands for accuracy and the degree of outcome dependence between the perceiver and the perceived (Erber & Fiske, 1984; Neuberg & Fiske, 1987). A key aspect of this theory is that there is a single continuum that accounts for the way we form impressions of other people.

Brewer and Harasty Feinstein (1999) argue instead that processing may take two different paths. On the one hand, the perceiver may initially categorize a target person and engage in category-based processing using a stereotype. When processed in this manner, the representation of the target will be stored in memory as an instance, even if a disconfirming one, of the category and the impression formed of the person will be based on those observed behaviors or inferred traits that are relevant to the category. Even in the case of individuation, the distinguishing information about the perceived individual will still be stored along with the initially invoked social category. On the other hand, person-based processing may occur. In this case, new information is processed without respect to category membership. This results in further differentiation of the person through a process of personalization and in this case information unrelated to the category as well as category-confirming and category-disconfirming information will be incorporated into the impression. Moreover, the representation in this case is independent of broader social categories. Brewer and Harasty Feinstein view category-based processing as a manifestation of intergroup perception whereas person-based processing occurs with interpersonal perception. The latter would be expected to result in a more differentiated, non-stereotypical view of a person.

The processes of individuation and personalization are expected to yield a more distinct representation of the perceived individual and more differentiated view of a group’s members. Although changes to the general stereotype may not result from such disconfirming cases due to subtyping (Brewer & Harasty Feinstein, 1999; Fiske & Neuberg, 1990; Kunda & Oleson, 1995; Weber & Crocker, 1983) or other processes, the formation of subgroups and individual exemplars in memory provide alternate representational structures for perceiving members of a social group. Moreover, the presentation of personalizing behavioral information has been shown to dominate predictions and inferences to the exclusion of stereotypes when the information is clearly diagnostic, and to weaken the effects of stereotypes in some cases even when the information is not directly diagnostic (see Kunda and Thagard, 1996 for a review).

These latter studies suggest that stereotypes may be applied primarily when little else is known about someone. However, there are challenges to this view. Brewer and Harasty Feinstein (1999) note that despite decades of close interpersonal contact between Serbs and Muslims, civil war and genocide was not prevented from occurring in Bosnia-Herzegovina. Another, and perhaps the most conspicuous, counter-example is the phenomenon of self-stereotyping. The fact that people will apply stereotypes to themselves implies that no amount of additional information is likely to eliminate categorization and stereotyping. In the typical case, each individual possesses more detailed information and a more differentiated view of the self than anyone else does. So, using the example of stereotype threat, why is a person unable to muster a wealth of knowledge and many possible exceptions that contradict applicability of the stereotype, especially given findings that the most threatened individuals are those with the most psychological investment, and presumably greatest achievement, in the particular domain (Aronson et al., 1999; Leyens et al., 2000; Pronin, Steele & Ross, 2004)? And why is it that observation of one’s own achievements or counter-stereotypical examples don’t get tagged as exceptions to the stereotype and establish the self as a contradictory subtype of it?

The distinction between category-based and person-based processing provides one path from this quandary. If the stereotype is extremely salient in the setting it may elicit category-based processing of the self rather than person-based processing. The stereotype, once activated and applied to the self, may then inhibit other possible categorizations (Hugenberg & Bodenhausen, 2004; Macrae, Bodenhausen & Milne, 1995) and the accessibility of other self-knowledge. Thus, a highly differentiated view of the self, and the presence of personalizing information in the setting, does not prelude the application of category-based processing to oneself. Category-based and person-based approaches appear to represent different modes of processing information that access different representations, or different aspects of a more general representation, in memory.