A REEXAMINATION OF THE ROLES OF TYPICALITY AND CATEGORY DOMINANCE IN VERIFYING CATEGORY MEMBERSHIP

Four partial replications of Chumbley's (1986) speeded category verification experiments were performed in order to test the generality of his finding that category dominance, but not typicality, uniquely accounted for highly significant amounts of category verification reaction time (RT) for both exemplar-category (E-C) and category-exemplar (C-E) orders. The multiple regression results for yes responses revealed that, contrary to Chumbley's findings, typicality produced significant effects in the 12 conditions used over the 4 experiments, whether C-E or E-C conditions. However, category dominance produced significant effects only in conditions using an E-C order, consistent with Loftus (1973). Familiarity was the strongest unique predictor of RT for no responses. These findings, in combination with Chumbley's, demonstrated that typicality is a likely, but not certain, unique predictor of category verification RT, and that the category verification task is extremely sensitive to a large number of task parameters.