Logistic Regression Models of the Safety of Large Trucks

Abstract

Statistics show that crashes involving large trucks are generally more severe than those involving other vehicles because of the size, weight, and speed differential between trucks and other vehicles. Given the critical position of trucking in the process of economic recovery and growth, the improvement of truck safety and the mitigation of any negative impacts on non-truck vehicles are urgent issues. Statistical models have been used universally to identify the contributing factors to crash severities and to estimate injury probabilities. These methodologies, albeit addressing different issues, may provide mixed results and estimates with varying degrees of accuracy. The primary objective of this research was to investigate the effects of key determinants of the severity of crashes involving large trucks and to explore the relationship between the determinants. The secondary objective was to provide insight on statistical applications by evaluating three logistic regression models: multinomial logistic, partial proportional odds (PPO), and mixed logistic (ML) models. The model results showed that the majority of the coefficient estimates were consistent across the models studied. A few exceptions included young drivers and the use of safety constraints; these factors were not statistically significant in the ML model. The goodness of fit and model predictive power indicated that the PPO model produced results that more closely resembled the observations.