Measurement error models on spatial network lattices: car crashes in Leeds

Road casualties represent the leading cause of death among young people worldwide, especially in poor and developing countries. This paper introduces a Bayesian hierarchical model to analyse car accidents on a network lattice that takes into account measurement error in spatial covariates. We exemplified the proposed approach analysing all car crashes that occurred in the road network of Leeds (UK) from 2011 to 2019. Our results show that omitting measurement error considerably worsens the fit of the model and attenuates the effects of spatial covariates.