Determinants of Response Times in the Semantic Verification Task

Multiple regression techniques were used to investigate the contribution of various factors to within and cross-category variations in the response times (RTs) obtained in a semantic verification task. Results showed the two most important factors to be typicality and familiarity. Typicality was found to have a significant unique within-category effect on positive RTs, but no effect on negative RTs. In contrast, familiarity had a significant unique effect on negative RTS but no within- category effect on positive RTs. However, familiarity was the most systematic predictor of cross-category variations in both positive and negative RTs. The only other factor with a significant predictive role was category dominance but, in contrast to typicality and familiarity, the dominance effects were limited to a specific presentation order of the instance and category names, and thus appeared to be strategy-based. The results are generally compatible with a probabilistic view of concept representation.