Logistic Model for Rating Urban Bicycle Route Safety

Abstract

In response to renewed appreciation of the benefits of bicycling to the environment and public health, public officials across the nation are working to establish bicycle routes. During the past two decades a number of methods have been endorsed for the selection of "suitable" bicycle routes. These methods are limited in that they do not explicitly address bicycle safety nor do they reflect urban conditions. The purpose of this research is to develop an objective bicycle route safety rating model based on injury severity. The model development was conducted with a logistic transformation of bicycle crash data from Jersey City, New Jersey, for the period 1997 to 2000. The resulting model meets a 90% confidence level by using various operational and physical factors (traffic volume, lane width, population density, highway classification, and presence of vertical grades, one-way streets, and truck routes) to predict the severity of an injury that would result from a motor vehicle crash that occurred at a specific location. The modeled rating of the bicycle route's relative safety is defined as the expected value of the predicted injury severity. This rating is founded on the premise that safe routes produce less-severe crashes than do unsafe routes. The contribution of this research goes beyond the model's predictive capacity in comparing the safety of alternative routes by providing planners with an understanding, derived from objective data, of the factors that add to the route's safety, factors that reduce safety, and factors that are irrelevant.