Institute of Transportation, MOTC

　　This study aims to investigate the potential risks for bicyclists. The Rasch model is used to measure bicyclists’ subjective perceived risk, and logistic regression is constructed to find contributor to the bicycle accidents casualty. Besides, we compare the Rasch and logistic model to discuss the gap between subjective and objective view of the potential risks. Our result also demonstrates some situations may be dangerous but cyclist did not aware of it which is high potential risk. Furthermore, we characterize bicyclist by DIF test and ANOVA shows that male and older bicyclist ride at night has a higher risk to casualty accident. The finding of this study can help government to build a more safety riding environment for bicyclists.