Driver Behavior as Predictor of Collision Risk using Ordered Probit Approach

Abstract

Increasing travel demand during the last decades has eventually increased the vehicle miles traveled (VMT) and motorization by road users, which increased the percentage of road traffic crashes (RTCs). Human factors are the most dominant factors causing highest percentage of RTCs across the globe. Among human factors, drivers’ faults are the most perilous ones because commencing slight mistakes by the drivers can be disastrous for all of the occupants and road users. Pakistan being a developing country suffers a huge social and economic loss each year in the form of road traffic mortalities and injuries. Also, most of the drivers’ population in the country is uneducated or low educated which worsens the situations. This study focuses on the investigation of factors affecting the involvement of drivers in RTCs. Because of ordinal nature of response variable i.e. accident involvement, an ordered probit model is extensively used. The estimation results predicted that drivers’ age, driving experience, hitting back object during reverse motion, reduce speed on wet or slippery roads and use of cell phone or Bluetooth during driving significantly affect the involvement of drivers in road crashes. The authors believe that the outcomes of this study will evolve a primary understanding of the transportation and traffic enforcement agencies in identifying risk factors of drivers regarding accident involvement and thus will mitigate the hazardous situation in the country. . Key Words: Driver Behavior Questionnaire, Road Traffic Crashes, Ordered Probit Model