Memory representation of motor skills in individuals with profound mental retardation
The purpose of the study was to examine whether individuals with profound mental retardation (PMR) create memory representations of a motor skill according to the predictions of a prototype model or an exemplar model. In Experiment 1, the predictions of the prototype and exemplar models were tested with regard to novel task transfer performance. Six male adults with PMR (CA: M = 44.75, SD = 12.17 years) threw beanbags to three distance exemplar during the acquisition, followed by four novel transfer exemplar distances, including two different and two similar exemplars. During 8 acquisition days participants performed 15 trials per exemplar. On day 9 participants performed 5 trials per acquisition exemplar and 15 trials per transfer exemplar. For constant (CE), absolute (AE), and variable (VE) errors, percentage scores were obtained by dividing the error scores by the exemplar distances. In Experiment 1, one-way repeated measures ANOVAs resulted in significant exemplar main effects for AE% and VE%, indicating that the four novel transfer exemplars were not performed equally well. However, there were no differences between the far and near exemplars for AE% and VE%, indicating that the degree of similarity between the transfer exemplars and the acquisition exemplars did not influence transfer performance. These results support the prototype view for memory representation of motor skill. In Experiment 2, the predictions of the prototype and exemplar models on 48-hr and l-week delayed retention tests were examined. Six male adults with PMR (CA: M = 40.67, SD = 5.65 years) practiced three exemplars of the beanbag throwing task and then performed two old and two novel exemplars for each retention test. One-way RM ANOVAs indicated that the participants maintained overall accuracy and consistency in performance across the retention intervals. Previous experience with an exemplar did not benefit retention performance. These results support prototype model predictions. Additional analyses indicated that individuals with PMR needed concurrent visual information of the target to perform the throwing task accurately and were able to generalize the beanbag throwing skill to the horseshoe throwing task.