Stereotyping based on apparently individuating information: Trait and global components of sex stereotypes under attention overload

Abstract

Although considerable research has shown that stereotypes are used to form impressions of others, other research shows that the effects of stereotypes can be attenuated or even eliminated by presentation of certain types of individuating information. Such information is usually in the form of traits or behaviors that imply traits. We hypothesized that sex stereotypic impressions can exist at both a specific trait level and a more global (stereotype fit) level, and that the degree of correspondence between the two components depends on the amount of attention paid to the target information. Subjects formed impressions of either a male or a female target who performed either several male-typed behaviors or several female-typed behaviors and several sex-neutral behaviors. The speed of presentation of the information was manipulated to create two levels of attentional processing load. Measures of sex stereotyping at both the trait and global levels were made. Subjects in both attention conditions appeared to stereotype the target based on his or her behaviors rather than on his or her sex, as their ratings of traits which were sex-typed but irrevelant to the behaviors were influenced by the sex-type of the behavior much more than by target sex. Whereas trait-level measures were responsive only to the behavioral information, the global-level measures were influenced both by target sex and by the sex-type of the behaviors. As predicted, the degree of correspondence between the two components depended on the amount of attention available for processing. When subjects had more time to attend to the information, their global and trait-level impressions were relatively independent. When subjects had little time to attend to the information, their trait impressions were driven mainly by their global impressions. Thus it appears that impressions are more likely to be stereotypic when people do not have time to integrate more individuating information, even when they make correct inferences from that information. The implications of these findings for the measurement of impressions and for stereotyping in naturalistic settings are discussed.