Varied and Fixed Error Limits in Automated Adaptive Skill Training.
Abstract
Three strategies of manipulating error limits were compared with a fixed error limit in an adaptive training system. Seventy-two (72) subjects were given training in a bidimensional, high-order, compensatory tracking task. The velocity/acceleration ratio of the control dynamics was the adaptive variable. Three of the four adaptive training groups required more time to attain the learning criterion than did the control group. A training group with increasingly rigorous error limits required more time to reach the exit criterion than did a group with the increasingly lax error limits. The changing response requirements of the adaptive task suggest that interference produced the inferior performance. No differences among groups were demonstrated in a retention test, and there was no transfer from the training task to a simulated flight control task for any experimental group. The model of adaptive training is seen as too simplistic to be productive. It is concluded that future research should be formulated in the context of a more comprehensive model of human learning. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1973
- Accession Number
- AD0764763
Entities
People
- William Henry Crooks
Organizations
- University of Illinois Urbana–Champaign