What Changes Occur During Complex Skill Acquisition?
Abstract
This paper presents and discusses the results of two separate but parallel studies of validity decrements of ability tests predicting skill acquisition and skilled performance. Hypotheses based on two explanations of the documented predictive validity decrements are tested. One explanation emphasizes transfer of training effects from learning and performing a complex task to the ability tests comprising similar elements of skills and knowledge. The second explanation emphasizes the effects of regression to the mean of ability measures in high ability groups that are selected, learned, and performed complex tasks. None of the hypotheses derived from these explanations for predictive validity decrements was supported. The results based on written pretest measures of ability replicated the basic phenomenon of decreasing predictive validities described in the literature Analyses of predictive validities of computerized tests did riot replicate the validity decrements across time and blocks of trials. Several analyses documented the advantages of exploiting assessments of performance taken during training as additions to prediction equations. The results of these analyses consistently showed that measures of performance obtained during training accounted for significant increments in variance beyond that accounted for by either pre- or post-training ability measures. The importance of this set of findings for practical, operational, solutions to the problems caused by predictive validity decrements is stressed.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1993
- Accession Number
- ADA262865
Entities
People
- Charles L. Hulin
- Kathy A. Hanisch
Organizations
- Iowa State University