Modeling Individual Difference Factors in a Complex Task Environment

Cognitive models are often used to predict the average performance of a population. For many purposes, however, generating predictions of individual performance is crucial. We propose a methodology in which the ACT-R architecture is extended, through the setting of architectural parameters that represent individual differences, into a model of individual behavior. This approach can provide a vast range of predictive and diagnostic capabilities from a modest initial investment of resources.