Abstract
The present experiment investigated the effect of varying the degree of task consistency on the performance and maintenance of skill in a semantic category visual search task. It is well established that for a wide variety of tasks, skill development is a function of the degree of task consistency. However, the effect of inconsistency on established skills has not been investigated to date. The present experiment included a consistent Training Phase, an Adjusted Consistency Phase, and a Retraining Phase. Subjects were trained for 6,000 Consistently Mapped (CM) trials on two different categories. Subjects then performed 4,000 trials in which one of the previously trained categories remained 100% consistent, while the other category became either 100, 67, 50, or 33% consistent. Task consistency was then restored and participants performed another 4,200 CM trials. The Retraining Phase included a New CM category. Results indicated that performance was disrupted by inconsistency, and that disruption increased as consistency decreased. Upon the return of task consistency, performance improved rapidly, although some performance disruption was still evident. The results are discussed in terms of visual search theories, and for their relation to training design.
Get full access to this article
View all access options for this article.