Severity analysis of single-vehicle left and right run-off-road crashes using a random parameter ordered logit model
Sunday Okafor, Emmanuel Kofi Adanu, Abhay Lidbe, Steven Jones
Abstract
OBJECTIVES: Single vehicle (SV) run-off-road crashes are a major cause of severe injury and fatality. Such crashes can result in different levels of severity depending on the direction (i.e., left or right) in which the vehicle runs off the road. This paper investigated the factors contributing to the crash severities of right run-off-road (R-ROR) and left run-off-road (L-ROR) SV crashes. METHODS: The study used SV crash data from the City of Charlotte, North Carolina, covering 2014 to 2017. Two separate random parameter ordered logit (RPOL) models were developed to estimate the contributing factors to R-ROR and L-ROR SV crash severities. The impact of the explanatory variables on the crash severity outcomes was quantified using the models' direct pseudo-elasticities. RESULTS: The model results showed that male drivers, Driving Under Influence (DUI), motorcycles, and dry road surfaces were significant contributing factors to R-ROR and L-ROR SV crash severities. Specifically for the R-ROR model, speeding, reckless driving, 1-2 lanes, and older drivers increased crash severity. For the L-ROR model, phone distraction, crossed centerline/median, 3-4 lanes, rain, and dark unlighted roadway increased crash severity. CONCLUSIONS: Based on the estimated parameters for the common significant variables in the two models, it was inferred that L-ROR SV crashes are more likely to result in severe crashes compared to R-ROR SV crashes. Hence, this study contributes to the literature on ROR SV crashes by providing additional insight into contextual factors influencing ROR crash severity for more effective countermeasures.