A hierarchical Bayesian spatiotemporal random parameters approach for alcohol/drug impaired-driving crash frequency analysis

•A Bayesian spatiotemporal approach is developed for crash frequency analysis.•This model can distinguish unobserved heterogeneity in spatial-temporal instability.•Spatiotemporal interactions are modeled using alcohol/drug impaired-related crash data.•The results indicate the appropriateness to integrate heterogeneity in the models.