Analysis of Factors Influencing Bicycle–Vehicle Interactions on Urban Roadways by Ordered Probit Regression

Abstract

A study of the impacts of roadway environment, motorist behavior, and bicyclist behavior on bicyclist–motorist interactions was based on video footage of traffic movements during peak commuting hours at four locations in Austin, Texas. The study developed three unique models of ordered probit regression that describe bicyclist lateral location, bicyclist–motorist interaction movement, and bicyclist–motorist lateral interaction distance. This structure of discrete choice model was used for the first time to address bicycle–vehicle interactions and offered more meaningful results because it used a latent measure of bicyclists’ and motorists’ mutual acceptance and comfort level sharing a roadway. It was demonstrated that adding “sharrows” and “Share the Road” signage promoted safer interactions on narrow urban roadways, whereas simply widening travel lanes and adding on-street parking did not necessarily improve bicyclist–motorist interactions.