Macroscopic spatial analysis of pedestrian and bicycle crashes.

摘要: Abstract This study investigates the effect of spatial correlation using a Bayesian framework to model pedestrian and bicycle crashes in Traffic Analysis Zones (TAZs). Aggregate models for were estimated as function variables related roadway characteristics, various demographic socio-economic factors. It was found that significant differences present between predictor sets crashes. The Poisson-lognormal accounting TAZs counties retained nine significantly different from zero at 95% credible interval. These – total length with 35 mph posted speed limit, number intersections per TAZ, median household income, dwelling units, log population square mile percentage households non-retired workers but auto, one long term parking cost, employment TAZ. A separate distinct set predictors crash model. In all cases performed better than did not account among TAZs. finding implies should be considered while modeling aggregate or macro-level.