The Left Turn Problem Using the Drift-Diffusion Model to analyze Bias and Speed-Accuracy Tradeoff when estimating Time-To-Contact.

Abstract

In daily life, it is essential to estimate the time needed for an object to reach a certain point, known as the time-to-contact (TTC). However, people are not always accurate in estimating this, as seen by the high percentage of traffic incidents when turning left with an oncoming vehicle at an intersection. To analyze which processes underly the estimation of the TTC, a drift-diffusion model (DDM) was fitted to the data of a perceptual choice task with manipulations. Also, gender differences were analyzed. In total, 100 participants performed the online task. The results show a change in the parameters of the DDM when a time-pressure is implemented, indicating a change in information processing. To compensate for the timepressure, participants relied more on a bias towards prior knowledge of distance. This bias was not evident when the participants were not rushed. Regarding gender differences, women tend to be more conservative in their TTC estimations. The study has enabled an innovative approach to analyze and explain traffic-related problems by combining the DDM with a manipulated perceptual choice task.