Time Perception: Beyond Simple Interval Estimation

Reasoning about and with time has never been studied extensively in the context of cognitive architectures. We present a temporal reasoning system that can make accurate predictions for two classic interval estimation experiments by Rakitin et al. (1998) and Penney et al. (2000). The system is implemented as an additional module for ACT-R. In combination with ACT-R’s other mechanisms, it makes accurate predictions about experiments that go beyond just estimating time. We demonstrate this by modeling data from a choice-reaction task by Grosjean et al. (2001).