Traffic conflicts have been used to define the potential for traffic accidents. However, conflicts defined by vehicle and pedestrian interactions have not produced reliable techniques to explain pedestrian/vehicle accidents. This study was conducted to determine the relationship between pedestrian/vehicle conflicts and accidents in order to develop a reliable model to predict the occurrence of pedestrian accidents. Accident group models were developed using discriminate analysis for the cities of Washington, D.C., and Seattle. Along with counting the conflicts that were used to define these accident groups, exposure measures such as pedestrian volume, vehicle volume, number of lanes, and type of traffic control aided in the explanation of pedestrian accident variance. Further research was recommended to investigate refined variable definitions along with the use of a larger accident data base.
Abstract