期刊名称：International Journal of Statistics and Applications
印刷版ISSN：2168-5193
电子版ISSN：2168-5215
出版年度：2018
卷号：8
期号：2
页码：35-41
DOI：10.5923/j.statistics.20180802.01
语种：English
出版社：Scientific & Academic Publishing Co.
摘要：Traffic crashes occur at different rates depending on several geometric conditions and actions of the drivers. Hills and curves have different crash rates and severities that level and straight segments of roadway. This paper presents data that were collected and analyzed using a binary probit model to determine the characteristics associated with higher crash severity of different grades (level, upgrade and downgrade) along with curve conditions (straight, turning left and turning right). Additionally, the paper continues the use of the binary probit model to analyze specific aspects that lead to run off the road crashes. The models developed in this research show that crash severity and run off the road crashes are often due to driver operating characteristics and not environmental or roadway factors. The application of the binary probit model demonstrates a convenient tool to analyze a large number of variables containing categorical crash data.
