Positional Mediation of Intraserial Associations

Abstract

The subjects learned to associate each of ten rectangles in a horizontal array with a different one of ten responses, the position cues for the responses being presented in an inconsistent serial order. The results of this spatial learning on intralist intrusions, on a subsequent test of associations, and then on a derived paired-associate task closely resembled the pattern of interitem associations typically produced by serial learning. It is concluded that the consistent temporal sequence in which items in a serial list are presented is not a necessary condition for the characteristic pattern of interitem associations that results from serial learning. A number of methods have been devised to measure the strength of the interitem associations that are assumed to be formed as a subject learns a serial list. Three of the most widely used methods are (a) the analysis of intralist intrusions, (b) the association method, and (c) the serial/ paired-associate transfer paradigm. For the analysis of intralist intrusions, each inappropriately placed response is classified by its displacement from its correct location in the serial list. The pattern typically emerging from such an analysis (e.g., Slamecka, 1964) is that when an intralist intrusion occurs, the response tends to be an item occupying a position in the serial list that is in close proximity to the position of the correct response. As the separation between two items in the list increases, the frequency with which each is erroneously emitted in place of the other decreases. Where the association method is used, each of the items of a previously