A separate analysis of crash frequency for the highways involving traffic hazards and involving no traffic hazards.

Considering the varying effects of influencing factors, this study aims to model the relationship between crash frequency and influencing factors for the highways involving traffic hazards and involving no traffic hazards separately. Two spatial-temporal interaction effect models are built with highway crash data occurring on the highway network comprising 11 highways managed by the China Communication Construction Co (CCCC) Asset Management Company. The model results show that there exist no significant spatial-temporal interaction effects for the highway crashes. The highway crash frequency for the highways involving traffic hazards is significantly larger than that for the highways involving no traffic hazards. The influencing factors such as the highway segment length, monthly traffic volume, and pavement width are found to exhibit much bigger effects for the highways involving traffic hazards than for the highways involving no traffic hazards. Results of this study provide supports for the argument that slightly different policies should be developed for the highways involving traffic hazards. For the highways involving traffic hazards, additional measures should be proposed to detect and repair destroyed road barriers as quickly as possible because this type of traffic hazard has a higher probability of causing highway crashes. [ABSTRACT FROM AUTHOR]