Validating Cognitive Human Models for Multiple Target Search Tasks with Variable Length Markov ''Chains

This paper describes a method we developed heuristically to validate cognitive models with the aid of variable length MARKOV chains. The focus of the proposed method is the validation of sequences of actions rather than the validation of time consumption. This method we applied in the investigation of visual attention in multiple target search tasks on computer-based radar displays. For the symbolic cognitive model description we used the visual module of the ACT-R/PM theory.