ABSTRACT
Introduction
The purpose of this study is to examine how reaction time (RT) distributions of an automobile driver change, in a stochastic manner, with the driver's arousal state. Most of the assessment indices proposed thus far (for example, headway time and time-to-collision) are computed solely from physical parameters, such as driving velocity and distance from a lead car (Evans, 1991; Hayward, 1972; Vogel, 2003). Although it is well known that human factors like arousal state are strongly related to automobile accidents (Dinges, 1995), human factors were not included in these indices. It is also well known that RTs during driving are strongly dependent on the driver's arousal state (Wlodarczyk et al., 2005), but how the RT distribution changes with the arousal level has yet to be determined. such stochastic properties of RTs are crucial for computing the probability of collision (POC), an index of the likelihood of an accident at a given moment (Matsuki et al., 2004).
