نوع مقاله : یادداشت پژوهشی
عنوان مقاله English
نویسندگان English
Pedestrians are the most vulnerable road users, and the expansion of exclusive bus lanes in developing megacities has created complex safety challenges at pedestrian conflict points. Despite the importance of exposure-based indicators in predicting crashes, prior research has not systematically compared these criteria within exclusive bus lane corridors, nor has it distinguished between crashes occurring inside and outside these lanes. This study aimed to identify the most effective exposure criterion for predicting pedestrian crashes and to examine unobserved heterogeneity at the neighborhood level.
Five years of crash data from 33 neighborhoods within the catchment area of exclusive bus lanes in Mashhad, comprising 396 spatial–temporal units, were analyzed. Four exposure criteria were tested: pedestrian crossing volume at grade-level crossings, average daily walking time, bus station passenger volume, and neighborhood population, alongside other predictors. Baseline negative binomial and Poisson models were estimated for crashes outside and inside the exclusive lane, respectively, together with their random-intercept counterparts. Models were compared using AIC, BIC, likelihood ratio tests, and marginal elasticities.
Results showed that pedestrian crossing volume was the best exposure criterion for crashes outside the exclusive lane, while station passenger volume was best for crashes inside the lane. Random-intercept models outperformed baseline models, with greater improvement for crashes outside the lane. Elasticity analysis revealed that crashes outside the lane are driven by accessibility and pedestrian environment factors, whereas crashes inside the lane are associated with corridor density. These findings confirm that the two crash environments follow distinct exposure mechanisms and require different targeted interventions. The results provide an empirical basis for selecting appropriate exposure indicators and prioritizing pedestrian safety measures in exclusive bus lane corridors according to crash environment type.
کلیدواژهها English