Pattern of disconfirming information and processing instructions as determinants of stereotype change.

This experiment examined the effects of pattern of disconfirming information (concentrated vs. dispersed) and processing instructions (focus on similarities vs. differences vs. control) on stereotype change. If subtyping and perceived typicality are central to the stereotype change process, then processing instructions designed to affect these processes should affect stereotyping. There was lower stereotyping when perceivers focused on similarities between group members, and after exposure to a dispersed pattern of disconfirming information. Only the main effect of pattern was mediated by the perceived typicality of disconfirmers, but not by an index of subtyping based on clustering of information from disconfirmers in recall. Results support a model of stereotype change in terms of the impact of disconfirming group members who are also seen as typical of the group; subtyping of extreme disconfirmers may work in parallel, or later, and contribute to the long-term maintenance of a stereotype