Proceedings of the Annual Meeting of the Cognitive Science Society

Anxiety plays an important role in predicting and avoiding future dangers, and adaptation under anxiety is assumed to be mediated by its influence on cognitive functions such as time perception. In this study, we constructed an ACT-R–based model that incorporates the amplification of anxiety over time. Using simulations of a turn-across oncoming traffic decision task in a driving context, anxiety was formalized not as a change in decision criteria but as a distortion of internal time progression. As a result, subjective time-to-collision (TTC) showed a logarithmic nonlinear relationship with objective TTC, and the between-condition differences corresponded to the difference structure of subjective time judgments observed in human experiments. Furthermore, examination of turn timing reproduced the anxiety-specific freezing process, in which the probability of performing the action decreased over time if it was not executed early.