Compare Between Logistic Regression and Dream methods on identifying risk factors in vehicle-pedestrian interactions

Statistics injured and killed pedestrians in recent years accidents, it is an expression of vulnerability of them. Identify risk factors interact is one of the important ways of understanding the causes of accidents. In this study, risk factors of interactions using logistic regression and Dream (analysis error and reliability, driving) methods, were compared with each other. Data were collected, using naturalistic driving studies (NDS) 29 driver bahavior at 289 interactions in the city of Babol during peak hours. The results showed that drivers and pedestrians risk behaviors, such as driving at high speed or running affect on interactions. The results of logistic regression due to statistical tests are accuracy and high reliability in identifying the risk factors of interaction, however, study in detail due to multiple diagrams is possible using the Dream method.