Mixed logit models for examining pedestrian injury severities at intersection and non-intersection locations

Abstract

Abstract Pedestrian injury has become a national traffic-safety concern as the share of pedestrian fatality continues to increase in the last decade. Pedestrian injury severities are influenced by many factors that include driver, pedestrian, vehicle, roadway, temporal, and environmental characteristics. Results indicate that some of the factors affecting pedestrian injury severity at intersection and non-intersection locations are statistically different and using the same model to perform the estimate at both locations may result in biased results. However, few studies have been conducted to explore different contributing factors at such locations. Mixed logit models are developed to independently identify the contributing factors to pedestrian injury severity resulting from crashes at intersections and non-intersections. The estimation shows factors such as male driver, alcohol, pedestrian above 65, truck, and higher speed limit significantly increase the probability of pedestrian serious injury severities in both locations. However, the impacts tend to be more severe at intersections. Urban and wet road surfaces decrease the likelihood of suffering fatal injury at intersections. Furthermore, crash time only has impacts at intersections, while traffic control, severe weather, and day-of-week only have impacts at non-intersections. The results provide insights on developing more effective countermeasures to promote pedestrian safety.