Understanding relationship between road median type and accident frequency

摘要

Road medians play an important role in the design of four or more lanes highways with high traffic volume and the large number of accidents. This study focused on effect of factors (such as traffic volume, road characteristics, land use and road median types) on the number of accidents. The data was collected from nationwide highway, and further used to build the models by using Generalized linear with Negative binomial model distribution classified by road median types including raised-, depressed-, flush- and barrier median. The result found that the increase in traffic volume and truck ratio are positively associated with more number of accidents in all four road median model. The finding also recommends that raised-, depressed- and flush median should be implemented within agriculture and rural area in order to decrease crashes frequency. The contributions of this study could be used as guideline regarding highways median type implementation decision-making to enhance road safety in Thailand.