Paired-Associate Analogue of Language Learning: Role of Reinforcement, Repetition, and Number of Rule Instances

Abstract

Adults learned 25 paired associates, forming a 5 by 5 matrix, in which the stimuli were 2-digit numbers and the responses were 2-letter pairs; 23 pairs (“regulars”) followed matrix rules; 2 pairs (“irregulars”) did not. on study trials, Ss were presented 21 regular pairs once and 2 irregular pairs twice; test trials were with all 25 stimuli. Ss learned the pairs more rapidly than those in a similar study which used fewer pairs and a smaller matrix. Ss receiving verbal reinforcement for correct responses on test trials performed no better than Ss not reinforced. Ss receiving both verbal reinforcement and repetition of correct responses on test trials learned fastest. Performance on irregular pairs was initially better than on regular pairs. Errors on the irregulars involved generalization of regular responses.