A counter model for implicit priming in perceptual word identification

Abstract

A model for the identification of briefly presented words is presented. The model accounts for data from naming and forced-choice experiments in which factors such as similarity of alternatives and stimulus presentation time are varied. The model assumes that counts are accumulated in counters that correspond to words and that a word is chosen as a response when the number of counts in its counter exceeds the maximum of the numbers of counts in other counters by a criterial value. Prior exposure to a word causes its counter to attract more counts than it otherwise would, and this yields priming effects. Ten experiments are presented, and the model provides correct predictions for the data. Implications of the model for research in implicit memory are considered.