Are more interactions at intersections related to more collisions for pedestrians? An empirical example in Quebec, Canada

This paper examines the relationship between reported collisions involving pedestrians and interaction rates based on field observation of street crossing behavior at selected urban intersections in Quebec, Canada (n=60). Data from a naturalistic observational study of pedestrian street crossing behavior was used to calculate interaction rates (n= 4286 observations) per intersection. Collisions (2003-2012, n=358 pedestrian collisions) were mapped and pooled at the same intersections to evaluate the relationship between the two. We found a positive and significant correlation between collisions (all-years) and interaction rates (Spearman’s coefficient between 0.329 and 0.340). This seems to be particularly the case on larger arterial roads.