TRAFFIC ACCIDENTS, ANALYSIS AND COST MODELING

Jordan is one of the most Arab countries with a high rate of traffic accidents resulting in the deaths of many people. More than seven hundred people per year die in Jordan because of car accidents and has been considered as the highest accident rate 12.8 per million of the population comparing with the World Health Organization (WHO) that have average rate of 10.3 per 100000 of the world population. Therefore, this study conducting analysis, evaluation, and prediction of traffic accident and traffic accident cost for ten years data. Different prediction statistical model were developed related to total number of fatalities, sever injuries, slight injuries, and property damages which are considered as dependent variables. Twelve independent variables were selected from 21 driver behavior mistakes as a highest ranking from previous studies. The results of analysis shows that five prediction model were developed for each dependent variable with high correlation coefficient (R2 ). The most independent variables to predict fatalities were using incorrect lane, priorities false for pedestrian &vehicles, and incorrect reversing. Other variables were significant on sever injuries, slight injuries, and property damage such as exceeding speed limit, tailgating, and not taking the necessary precautions while driving. According to prediction of accident cost anther five models were developed for each dependent variable with high (R2 ). The most independent variables have an impact on accident cost are using (incorrect lane, priorities false for pedestrian &vehicles, and incorrect reversing.