The optimal use of simulation.

Abstract

Abstract : One hundred subjects were run on a foot-hand-eye coordination task. The apparatus used was a collator. It was set up as a simulator as well as the 'real' task. Twenty subjects (10 male and 10 female) were run in each of five conditions. Condition I - training on collating three pages then transferred to collating eight pages. Condition IA - trained on three pages then overtrained before transferring to eight pages. Condition II - trained on five pages then transferred to eight. Condition IIA - trained on five pages then overtrained before transferring to eight. Condition III - trained immediately on eight pages. For the males, overtraining degraded performance on the transfer trials. This was not true for the females. The closer the training was to the 'real' task (8 pagges) the poorer the performance of the males on the transfer performance. Overtraining males who reached criterion early in the initial training had the greater deleterious effect on the transfer trials when compared to those reaching criterion later. This was not true for the females. For the males, no prior training produced the best terminal performance on the transfer task. This was not true for the females. Some suggested explanations of the sex difference are given and some hypothesis are made concerning the optimal use of simulation training. (Author)