Journal of Transportation Engineering · 2013 · 103 citations · 33 references
Public HealthInjury PreventionCausal InferenceTransportationVehicle Safety EngineeringAccident PreventionInjury EpidemiologyInjury RiskTraffic Injury PreventionTraffic SafetyTransportation SafetyAccident Severity ModelingCrash Injury SeverityCrash Severity AnalysisEconomic Impact PredictionPotential Economic ImpactTraffic Accident AnalysisUnobserved ConfoundingUnobserved Heterogeneity
In recent years, a growing concern related to large truck accidents has increased owing to the level of injury severity that can be sustained and to the related potential economic impact. Current studies related to large truck–involved crashes are scarce and do not address the human factors that can greatly influence accident outcomes. This study presents an analysis of data from the fusion of several national data sets addressing injury severity related to large truck–involved crashes. This is accomplished by considering human, road environment, and vehicular factors in large truck–involved crashes on U.S. interstates. A random-parameter ordered-probit model was estimated to predict the likelihood of five injury severity outcomes—fatality, incapacitating, nonincapacitating, possible injury, and no injury. The modeling approach accounts for possible unobserved effects relating to human, vehicular, and road environment factors not present in the data. Estimation findings indicate that the level of injury severity is highly influenced by a number of complex interactions between factors, and the effects of some factors can vary across observations.
33
Mixed MNL models for discrete response
Daniel McFadden, Kenneth Train · Journal of Applied Econometrics · 2000 · 4.1K citations