2020 Volume 6 Issue 2 Pages A_31-A_40
This paper analyzed relationships between driver’s anxiety and unsafe driving tendency in order to enable risk assessment of potential traffic accidents. Using an indoor driving simulator, we collected the vehicle behavior data. Additionally, we collected levels of anxiety by interviewing the examinees after the completion of every test run. The examinees were classified via cluster analysis (Ward’s method) using their level of anxiety. It was found that, when the examinees were classified into three clusters, the classification was the easiest to interpret. So we focused on the examinees who changed their level of anxiety drastically between the different course conditions. Based on the results of the analysis of variance of jerk and distance from the center of the lane, we established that there are positive relationships between anxiety and unsafe driving tendency. Furthermore, on the basis of analysis of maximum acceleration, we confirmed that unsafe driving tendency tended to be the source of unsafe behavior, though this tendency was only partially verified.