Injury Severity Analysis in Taxi Crashes: A Random Parameters Approach with Heterogeneity in Means

The purpose of this paper was to explore the risk factors that affected crash severity when casualties involved taxis. Using 23,436 taxi crash data recorded by police from 2009 to 2019 in Hong Kong, a random parameter logit model with heterogeneity in means was established to explore the relationship between crash severity and explanatory variables, which outperformed the random parameters model only in terms of goodness-of-fit. The results indicated that time of day, the number of vehicles involved, speed limit, traffic congestion, intersection type, the vehicle operation, the vehicle age, and sideswipe/rollover had significant effects on the injury severity, among which 17:00–20:59, severe traffic congestion, moderate traffic congestion, special intersection, and vehicle with the usage of 6–10 years had heterogeneous effects. The present study provided a theoretical basis and a sound toolkit for improving the traffic safety of taxis.