Organization in Free Recall Learning: Output Contiguity and Interresponse Times as a Function of Presentation Structure.

Abstract : The experiment was designed to validate the use of interresponse times (IRTs) as an index of organization in learning, and to determine whether the effects of input contiguity on organization and recall are due to direct representations of contiguity relationships in S's functional organization, or to the indirect potentiating effect of contiguity on item-dependent organization. Presentation orders and semantic relatedness among words were varied factorially in a multitrial free recall task. Words were either members of a superordinate category, or unrelated. Members of groups were either blocked or randomly dispersed on individual trials, and presentation orders were either constant or varied across trials. The results indicated that IRTs provide similar and supplementary information about organization when compared with indices of clustering and subjective organization. Differences in IRTs between members of different categories as compared to members of the same category were related to the degree of clustering according to those categories. The IRT between two words was related to the number of previous recall trials on which the pair occurred together as well as to the probability of adjacent occurrence of the pair on subsequent trials. On Trial 1, an interaction of contiguity and semantic relationships was obtained with clustering scores, but not with recall scores. In the multitrial situation, intratrial contiguity was effective for related words, and intertrial contiguity consistency was effective for unrelated words. There was no evidence for direct contiguity effects. Other effects of multiple potential bases for interitem relationships were considered.