作者: Zhenning Li , Xiaofeng Chen , Yusheng Ci , Cong Chen , Guohui Zhang
DOI: 10.1016/J.AMAR.2019.01.002
关键词: Major injury 、 Crash 、 Computer science 、 Random parameters 、 Unobservable 、 Bayesian probability 、 Impaired driving 、 Statistics 、 Crash frequency 、 Spatial heterogeneity
摘要: 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 time instability, severities, etc. However, only very limited body of research is dedicated to distinguish simultaneously address all these sources unobserved heterogeneity. In this study, hierarchical Bayesian random parameters models with various spatiotemporal interactions are developed issue. Selected for analysis the yearly county-level alcohol/drug impaired-driving related counts data three different injury severities minor injury, major fatal Idaho 2010 2015. The variables, daily vehicle miles traveled (DVMT), proportion male (MALE), unemployment rate (UR), percentage drivers 25 years older bachelor's degree or higher (BD), found have significant impacts on be normally distributed certain severities. Significant temporal spatial heterogenous effects also detected These empirical results support incorporation heterogeneity models.