Analysis of head-on crash injury severity using a partial proportional odds model

Abstract

Head-on crashes are one of the most severe crash types and always result in injuries or fatalities. To prevent and mitigate head-on crashes, factors that significantly affect the injury severity of head-on crashes must be identified before appropriate countermeasures can be explored. To bridge the gap between ordered and unordered response modeling, in this research, a partial proportional odds model is developed to analyze the factors that influence the injury severity of head-on crashes. The analyses are performed based on the data collected from Highway Safety Information System (HSIS) from 2005 to 2013 in North Carolina. The results of this research demonstrate that there are 14 factors that have significant effects on the injury severity of head-on crashes. Among them, the roadway with speed limit over 50 mph is found to increase the fatal crash most. Appropriate countermeasures are recommended according to the influencing factors that are identified. The model performance is also compared with ordered logit model and multinomial logit model. The partial proportional odds model can provide adequate fit without potential loss of prediction accuracy.