Modeling Fault among Bicyclists and Drivers Involved in Collisions in Hawaii, 1986–1991

Abstract

Subsequent to a review of trends in collisions between bicyclists and motorists in Hawaii during the period 1986 to 1991, characteristics of bicyclists and drivers involved in crashes are compared. On the basis of police-reported crash data it can be concluded that bicyclists tend to be young, male, and, not surprisingly, more likely to be seriously injured than motorists in bicycle–motor-vehicle collisions. Bicyclists are much less likely to be attributed with inattention than motorists, and slightly less likely to be attributed with misjudgment or alcohol or drug use than motorists. Bicyclists, however, are much more likely than motorists to disregard traffic controls or go the wrong way on a street just before becoming involved in a collision. Motorists are more likely than bicyclists to fail to yield, to engage in improper overtaking, or to follow too closely before becoming involved in a collision. The largest proportion of bicycle collisions occurs during the period 3:30 to 6:30 p.m. Other temporal and locational features of bicycle collisions are described. Among the most significant findings presented is that motorists in Hawaii are much more likely to be classified at fault than bicyclists. Whereas motorists are at fault in approximately 83.5 percent of incidents, bicyclists are at fault in only 16.5 percent of incidents. A logistic model is developed and used to explain the likelihood of motorists being at fault in collisions with bicyclists. Covariates that increase the likelihood of motorist fault include motorist age (squared), bicyclist age, bicyclist helmet use, and motorist turning actions. Variables that decrease the likelihood of motorist fault include motorist age, bicyclist age (squared), bicyclist alcohol use, bicyclists making turning actions, and rural locations.