Modeling Pedestrian Compliance Behavior at an Urban Signalized Intersection

抄録

Traffic accidents involving pedestrians are the major causes of fatality and injury at urban signalized intersections in India. This study investigates the effect of various factors on pedestrian compliance at urban signalized intersection. The data is collected using video recording and questionnaire survey in New Delhi, India. Chi-square and t-test are performed to identify statistically significant variables. Single factor model is developed using odds ratio (OR) statistics. Pedestrian compliance behavior models are developed using binomial logistic regression. The models have good predictive capabilities and their success rate of prediction is more than 80%. The data shows that the crosswalk compliance is only 16.4%. The major reason for non compliance is to save time and personal convenience. The findings from this study would help city planners, traffic engineers and policy makers to improve pedestrian safety at signalized intersection.