Stimulus recall following paired-associate learning.

Abstract

Subjects (Ss) learned a list of 7 paired associates. Immediately after learning they were presented the response items, 1 at a time, and were instructed to give the stimulus items associated with them during learning. The basic question was whether or not S could give the stimuli correctly. Does S learn an R-S association when instructed only to learn the S-R association? Four groups of 30 Ss each were used. Each group learned a different list. The 4 lists made up a 2 x 2 design in which stimulus similarity and response similarity were both high and low. For each list the stimulus items were nonsense syllables and the response items adjectives. The results show: in terms of syllables correct, recall of the stimuli (R-S) was 50% correct. In terms of correct positions the recall was 61%. While both stimulus and response similarity markedly affected rate of learning the S-R associations, only stimulus similarity produced significant differences on recall. R-S recall was not strongly related to strength of S-R associations; when response similarity was high, there was virtually no relationship. The results establish the existence of a phenomenon not hitherto reported but probably long suspected to exist. Whether or not it is a form of incidental learning requires further analyses.