Integrating Individual Differences Metrics into Cognitive Models

Abstract

Individuals vary in susceptibility to fatigue, resulting in different performance given the same stressors. Current performance modeling capabilities for fatigue typically ignore these individual variations, decreasing predictive accuracy. The current effort examines the predictive utility of integrating subjective individual differences metrics into cognitive models of cognitive processes crucial for operational performance. Stable individual differences metrics—circadian typology, typical sleep duration and sleep need, and hardiness—were measured via self-report questionnaires, and cognitive performance for psychomotor vigilance was measured during a 24-hr simulated air mobility mission where 39 pilots experienced restricted sleep opportunities while performing various mission tasks. The psychomotor vigilance cognitive model was developed in the Adaptive Control of Thought-Rational (ACT-R) cognitive architecture and paired with the ACT-R fatigue module. Individual differences metrics were integrated into the model via modulations of the fatigue module parameter :fpbmc through a Metropolis-Hasting procedure. Integrating individual differences into the cognitive model improved the predictive fit for each pilot, but the practical benefit of the individuation was mostly seen in a few participants. These results suggest that organizations may be able to increase the accuracy of fatigue models by incorporating stable individual differences information from surveys or other means, enhancing fatigue risk management.