Exemplars, Prototypes, and Similarity Rules

One of the major themes in categorization research during the past two decades has involved comparisons and contrasts between prototype and exemplar models (e.g., Estes, 1986 a; Hintzman, 1986; Homa & Chambliss, 1975; Medin & Schaffer, 1978; Nosofsky, 1987; Posner & Keele, 1970; Reed, 1972). According to prototype models, the observer forms an abstract summary representation of a category, usually assumed to be the central tendency of the category distribution. According to exemplar models, the observer stores the individual training exemplars of a category in memory. In both prototype and exemplar models, classification decisions are based on the similarity of an item to the underlying category representation.