Deciphering the Cognitive Complexity of Road Crossing Decisions: A Novel Model Integrating Time-Varying Drift Rates, Signal Suppression, and Utility Maximization

Road-crossing decisions exhibit unique characteristics not typically seen in classic decision-making tasks. To explore this complexity, we introduce a new time-varying decision-diffusion model to discover the cognitive processes behind how pedestrians assess an oncoming vehicle when preparing to cross a road. The model proposes a plausible cognitive mechanism where perceptual information relates to collision risk dynamically. We conducted two experiments, recorded behavioral and electroencephalogram (EEG) data, and applied the model to the data set. Our model effectively captures the intricacies of the decision-making process, revealing a distinct bimodal distribution of response times tied to safety-related events. The model closely aligns with the data, except in infrequent crossing events. Our EEG data show how the decision variable changes over time, with signal suppression leading to symmetric response time distributions. Together with the modeling result, we concluded that decision rate and collision risk play vital roles, mediated by three fundamental mechanisms: the time-varying drift rate, signal suppression, and utility maximization. We discuss how these elements converge to provide a testable framework explaining how pedestrians determine the safe moment to cross in front of an approaching vehicle.