Information Processing Components and Knowledge Representations: An Individual Differences Approach to Modeling Pilot Judgment

Abstract

The study described here represents another step in an ongoing program of research (cf. Wickens, Stokes, Barnett, & Davis, 1987; Wickens, Stokes, Barnett, & Hyman, 1988). Specifically, the present study contrasts high-time and low-time pilot judgment performance, using information processing components and knowledge representations in long term memory (LTM) as individual difference measures to predict performance. The objective was to determine which of these two classes of measures predicted pilot judgment performance for groups of varying levels of experience. Thirty pilots (15 high-time and 15 low-time) completed a cognitive abilities assessment battery. This battery was comprised of two components: domain-independent information processing measures, and measures of domain-specific knowledge representations. These pilots then flew a cross-country flight on MIDIS, a microcomputer-based decision flight simulator. Each pilot's performance was assessed on a number of in-flight decisions. No significant differences in absolute performance level were observed between high-time and low-time pilots, however the pattern of ability measures that predicted low-time pilot performance was different from those that predicted high-time performance. Specifically, high-time pilot performance was better-predicted by measures of domain-specific knowledge representations. Differences in predictor patterns suggest qualitative differences in decision-making strategies used by the two cohorts.