Learning Priorities in a Paired-Associate Task with Polygons and Numbers as a Function of Polygon-Set Variability

Abstract

Ss performed a paired-associate task with polygons and numbers, where polygon sets satisfied four different combinations of feature variability. Two groups of Ss working with two sets where forms were highly variable on the Jaggedness feature (HH, HL) consistently selected certain polygons for early pairing. These polygons were assumed to have high primacy, and their selection was independent of all feature measures of forms and their derivatives. Two groups of Ss working with two sets where forms were not highly variable on the Jaggedness feature (LL, LH) formed pairings at a slower rate. Within these groups, selection of individual polygons for early learning was dependent on several measures reflecting their extremity on both Jaggedness and Area measures. No such differences were obvious with groups differing in Area variability. Results are interpreted in terms of a hierarchy of response bases for early pairing where “Gestalts” are used in conditions of high Jaggedness variability and “extremes” in conditions of low Jaggedness variability.