Evaluation of effects of risk factors on injury severity of single-vehicle accidents

Approximately 30% of total road traffic accidents in India are single-vehicle accidents (SVAs). The main objective of the present study is to evaluate the individual and interacted effects of risk factors on SVA severity. Using ten-year (2011-2020) accident data reported from Itanagar and Shillong, India, two sets of binary logistic regression models (model-without-interaction and model-with-interaction) were developed. The present study considered two-way and three-way interaction analysis in the model-with-interaction. Modelling results show that in the model-without-interaction, 7 categorical variables (18-24 age group, 25-40 age group, above 40 age group, heavy motor vehicle, 6 PM-12 AM, rushed driving, and rollover) in the Itanagar study area and 4 categorical variables (female, light motor vehicle, out-of-control, and Run-off-Road) in Shillong study area were shown statistically significant for the SVA severity. For the model-with-interaction, there were 12 significant interactions (7 two-way interactions and 5 three-way interactions) in the Itanagar study area and 7 significant interactions (5 two-way interactions and 2 three-way interactions) in the Shillong study area. These findings will be useful for transportation policy-makers while implementing correctives to improve road safety in these regions.