Factors influencing road accidents in Sri Lanka: a logistic regression approach

Abstract

Road accidents are a major socio-economic problem in Sri Lanka. This study aims to identify the factors that mainly contribute to accident severity in Sri Lanka and to identify the significance of the factors for formulating the model. For this purpose, road accident yearly data were collected for the period of 2010 to 2014 from the Police Traffic Headquarters in Colombo and 13 factors were considered. In this study, Binomial Logistic Regression has been used to analyze the data due to the dichotomous nature of the dependent variable.Based on the results, concluded that variables such as light condition, age of the driver, the validity of the license, urban / rural, weather, vehicle type and age of the vehicle have a decreasing effect on the probability of a fatal accident. Similarly, remaining variables such as location type, alcohol test and accident cause have an increasing effect on the probability of a fatal accident. Among them, Accident Cause is the most important variable in the model.