Abstract
A token training programme on a shopping skills task was given to 48 institutionalized learning disabled adults, 15 subjects were the control group. The factorial design involved the following independent variables: Two training locations: treatment room in hospital; a small village shop. Two token delivery schedules: continuous; intermittent. Two methods of instruction on the training task: verbal instructions only; both verbal instructions and modelling the task by the trainer. The aim of the design was to facilitate the transfer to a supermarket and retention for three weeks after training, the shopping skills performance. HYPOTHESIS: (1) The predicted transfer promoting variables are: (i) training in the village shop. (ii) providing continuous tokens. (iii) providing verbal instructions only on training task. (2) The predicted retention promoting variables are: (i) training in hospital treatment room. (ii) providing intermittent tokens. (iii) providing both verbal instructions and modelling the task by the trainer. (3) Four of the 3-way variable interactions were predicted to maximise transfer performance, while the remaining four interactions were predicted to maximise retention performance. The analysis of variance was used for data analysis. RESULTS: Training session: (A) Performances improved across the four training trials. (B) Four training conditions were more effective than others. Transfer and Retention: (A) Trained subjects showed highly superior means of transfer and retention to those of the control subjects. (B) Similarly high transfer and retention means were produced by the eight training conditions. (C) Higher retention means were produced by subjects who: (i) trained in the village shop. (ii) received both verbal instructions and modelling the task by the trainer. (D) Transfer versus retention performances: (i) Control group showed significant decline of retention means. (ii) Trained subjects showed similarly high means on both tests. (E) Effects of a few non-therapeutic independent variables were significant as follows: Better performances were produced by subjects of: (i) Higher I. Q. (ii) Older age (both during training only) (iii) Better educational abilities during training and both tests. Results were discussed in light of the hypothesis, followed by recommendations for future application.