Cyclist injury severity analysis with mixed-logit models at intersections and nonintersection locations

Cyclist injury severity in bicycle-vehicle crashes is influenced by numerous factors that are related to driver, vehicle, cyclist, crash, roadway, temporal, and environmental characteristics. Results indicate that differences remain between cyclist injury severities sustained in crashes occurred at different locations (intersections and nonintersection locations). However, few research efforts have been made to examine and compare different contributing factors to cyclist injury severity at such locations. Therefore, this study aims to identify and analyze the impact of various factors on cyclist injury severity at intersection and nonintersection locations. Police-reported data from 2007 to 2014 in North Carolina are used in this study. Three mixed-logit models are developed based on the preestimated multinomial logit models. Market segmentation is conducted to analyze cyclist injury severity at separate locations. An ordered probit model is also developed and compared with the mixed-logit model at both locations. Furthermore, cyclists drinking alcohol; driving van, bus, or single-unit truck; motorists' fault; inclement weather; dusk or dawn are found to have a significant impact on cyclist injury severity at intersections. Cyclist gender, drivers drinking alcohol, vehicle speed, speeding, rural or urban areas, traffic control, curved road are found to have a significant impact at nonintersection locations.

Automated summary

The severity of cyclist injuries in bicycle-vehicle crashes is influenced by various factors related to different locations and contributing factors have significant impacts on cyclist injury severity at intersections and nonintersection locations. Factors such as alcohol consumption by cyclists and motorists, vehicle type, fault attribution, weather conditions, and time of day have been found to have significant impacts on cyclist injury severity.