Modeling Pedestrian Temporal Violations at Signalized Crosswalks: A Random Intercept Parametric Survival Model

This study aims to model pedestrian temporal violation behavior at signalized crosswalks. Video data of crossing were collected from three locations in China and used investigate the effect several factors on behavior. The was analyzed using relationship between waiting duration their endurance probabilities. A fully parametric with Weibull distribution behavior, cluster-specific heterogeneity among sites accounted for random intercepts. Six variables identified significantly affect behavior: gender phone distraction status, location type, volume, day week, time day. results show that pedestrians are likely disobey traffic regulations when there longer durations. Male have a higher tendency than females. Pedestrians distracted by phones durations undistracted pedestrians. Signalized road segment crosswalks associated propensity intersection more commit violations densities. Weekdays shorter weekends. violate morning midday evening. These findings give insights into better accommodate improve safety.