Logistic Model of Hit and Run Crashes in Calgary

Résumé

Hit-and-run cashes refer to traffic collisions in which at least one driver flees from crash scene without reporting the crash. In the City of Calgary, for example, they accounted for 18 percent of total traffic collisions in 2005. The objective of this study is to identify the environment and road characteristics that contribute to the occurrence of hit-and-run crashes in the City of Calgary. A logistic regression model was developed to delineate the likelihood of hit-and-run crashes as opposed to non hit-and-run crashes. Our study showed that compared to weekday and daytime collisions, weekend and night time collisions have significantly higher likelihood of hit-and-run. In terms of weather condition, clear weather exhibited the greatest chance of hit-and-run when compared to any other weather conditions. Moreover, hit-and-run crashes are quite likely to occur on undivided one-way roads and the roads with artificial light. As for driver related factors, female drivers aged at 55 or above showed the greatest likelihood as compared to other age groups. Based on the findings from this study, a set of countermeasures will be proposed in this paper.