Physical Task Performance: Complexity of the Ability-Performance Interface.
Abstract
Simple, general models may provide useful representations of physical abilities and task performance. Physical abilities can be represented by 3 to 6 general dimensions summarizing individual differences on a range of specific physical ability tests (e.g., Hogan, 1991; Itrers, Gebhardt, Crump, & Fleishman, 1993). Similar simplicity may suffice to represent the physical task domain. When people perform two or more physically demanding tasks, individual differences in task performance are moderately to strongly positively correlated (Arnold, Rauschenberger, Soubel, & Guion, 1982; Beckett & Hodgdon, 1987; Robertson & Trent, 1985). This correlation pattern suggests the existence of one or more general dimensions underlying differences in performance on the specific tasks. With these observations as a starting point, this paper examines the utility of general dimensions as summary measures to characterize task performance and to describe physical ability-task performance relationships. The potential value of modeling ability-performance relationships in terms of general dimensions can be illustrated by comparison to more common approaches to modeling physical task performance. Standard practice for predicting physical task performance involves measuring performance on one or more tasks and administering a battery of physical ability measures. Procedures such as stepwise regression then are applied to select a set of ability measures to predict each task. The result is one predictive equation per criterion with different predictors and/or predictor weights for each criterion. The set of regression equations defines the performance prediction model for the task set.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 05, 1995
- Accession Number
- ADA308696
Entities
People
- Ross R. Vickers
Organizations
- Naval Health Research Center