Modeling Automaticity and Strategy Selection in Dynamic Visual Detection

A model of human automaticity in a dynamic visual task (RADAR) is reported. This model represents a first approximation for a model that reproduces the main results from human participants collected by Young, Healy, Gonzalez, and Bourne (2007). The initial results on response time demonstrate high degree of fidelity (R2=0.98) using a process model derived directly from task requirements. The model leverages the ACT-R cognitive architecture (Anderson & Lebiere, 1998) and depends primarily on assumptions that have been imported from previous modeling efforts. The model is an instantiation of a dual-process theory of automaticity that is mediated by a strategy selection stage.