Abstract versus exemplar memory representations: the role of contextual interference

Abstract

The two experiments discussed were developed from a aphics. study, conducted by Chamberlin and Magill (1992b), which contrasted predictions from the schema abstraction model with those from the exemplar-based model. Recently schema theory has been used as a model to help explain the representation of motor skills (Schmidt, 1975); however, very few studies have addressed this issue. The current experiments were designed to test whether varied practice schedules would effect the development of memory representations, and to try and replicate the findings of Chamberlin and Magill (1992b). Both experiments required the participants to perform 297 trials per day, of three separate tasks, for four consecutive days, followed by a retention and transfer test on the fifth day. The three tasks either varied in the total distance moved (Experiment 1), or in the overall goal time (Experiment 2). All tasks required the participants to begin at the start location, move 450 upward to the right, then move 450 to the left and complete the movement with a 900 upward movement. During transfer there were four tasks, two of which were similar to a trained task, and two that were equidistant from two trained tasks. The results were similar for both experiments providing support for the schema abstraction model, and replicating those of Chamberlin and Magill (1992b). The error recorded during retention was similar to that during transfer', in addition there were no large differences between the transfer to near and far movements. A typical contextual interference effect was found during retention and transfer, as random participants outperformed blocked in Experiment 1 and in Experiment 2. These overall findings of the two experiments give support to the abstraction notion whether using the manipulation of overall duration, or overall movement distance.