Academic journal article Human Factors

Cognitive Ability Determinants of Elite Pilot Performance

Academic journal article Human Factors

Cognitive Ability Determinants of Elite Pilot Performance

Article excerpt


The concept of situational awareness (SA) has excited much interest and debate, as reflected in a recent special issue of Human Factors (Gilson, 1995). Whereas some (e.g., Endsley, 1995) see SA as an explanatory construct distinct from decision making and action, others (e.g., Sarter & Woods, 1995) view SA as a convenient label for various cognitive processes that must be involved in human interaction with complex, dynamic environments. By definition, such environments present the operator with multiple tasks that must be monitored, prioritized, and managed (Adams, Tenney, & Pew, 1995). Several authors (e.g., Endsley, 1995; Gaba, Howard, & Small, 1995) have raised the critical question of whether there are identifiable individual differences in the ability to acquire and maintain SA. Such differences might be attributable to differences in an inherent SA ability, attention management skills, or general cognitive ability.

Endsley (1995) and Endsley and Bolstad (1994) suggested that individual differences in SA could arise from underlying individual differences in specific abilities such as working memory capacity, perceptual speed, pattern-matching ability, cognitive complexity, mental simulation, and attention sharing. Numerous tests have been developed to measure these abilities, often for the purpose of pilot selection, and a number of batteries such as MICROPAT (Bartram, 1987) and BAT (Caretta, 1989) have been validated on measures of pilot performance during training. In general, such batteries provide moderate predictive validities for initial flying training success and lower predictive validities for measures based on advanced performance (Roscoe & North, 1980).

In contrast, Hopkin (1993) viewed SA as a higher-order attention management ability relatively independent of the elemental abilities that are measured by conventional batteries. Indeed, tests of attention management abilities have shown particular promise in predicting performance in environments requiring high levels of SA. Gopher and Kahneman (1971) showed that errors on a dichotic listening task were negatively correlated with initial flight training success. Gopher (1982) reported a more extensive validation using 2000 candidates for flight training in the Israeli Air Force.

Measures of selective attention were found to be predictive of subsequent pilot performance and to have low correlations with existing measures. Most predictive were errors in switching attention from one information channel to another. More recently, Tham and Kramer (1994) found that flight instructors made fewer switching errors on a dichotic listening task and showed greater ability to inhibit irrelevant stimuli in a lexical target/distractor task than did student pilots. Individual differences in another aspect of attention management - the ability to time-share between tasks - have also been shown to be predictive of pilot performance.

Damos (1972) initially showed the value of cross-adaptive measures of residual attention as predictors of student pilot performance. In that study participants' performance on a compensatory tracking task and reaction time on a digit-canceling task were used as measures of residual attention. Attention measures were correlated with rated performance in flight checks after 10, 20, and 30 h of flight training. The measures correlated even more highly (r = .69) with performance at 30 than at 10 h (r = .56). A subsidiary finding of interest was that the student pilots tested in the first study, who were drawn from the professional pilot program at the University of Illinois, had better scores on the measures of residual attention (an early name for SA) than did the second group, drawn from a general group taking basic flight courses.

Gopher and North (1974) developed an approach to the measurement of individual differences in attention allocation. Their participants performed the one-dimensional compensatory tracking task and digit canceling task used by Damos (1972) under different conditions of demand: separately, equal priority given to both tasks, and varying priorities given to either task. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.