LAUSR

Abstract Unobserved heterogeneity, which has been recognized as a critical issue in crash frequency modelling, generates from multiple sources, including observable and unobservable factors, space and time instability, crash severities, etc. However, only a very limited body of research is dedicated to distinguish and simultaneously address all these sources of unobserved heterogeneity. In this study, hierarchical Bayesian random parameters models with various spatiotemporal interactions are developed to address this issue. Selected for analysis are the yearly county-level alcohol/drug impaired-driving related crash counts data of three different injury severities including minor injury, major injury, and fatal injury in Idaho from 2010 to 2015. The variables, including daily vehicle miles traveled (DVMT), the proportion of male (MALE), unemployment rate (UR), and the percentage of drivers of 25 years and older with a bachelor's degree or higher (BD), are found to have significant impacts on crash frequency and be normally distributed in certain crash severities. Significant temporal and spatial heterogenous effects are also detected in all three crash severities. These empirical results support the incorporation of temporal and spatial heterogeneity in random parameters models.