The aim of this study is to investigate vehicle lateral displacement while driving on curved and straight rural road sections on an actual road environment and in a fixed-base driving simulator. Driver behavior in terms of lateral position and general vehicle path on an actual road has not been investigated until recently, mainly for the practical reason that few instruments could measure lateral position with sufficient accuracy, compared with speed measurement instruments, which were developed a long time ago. For example, the first induction loops were operated in 1928 in Baltimore and worked with sound (Fraser, 1984). Until recently the two main instruments used for measuring lateral position have been video analysis systems (Ashworth, 1976; Dickinson & Waterfall, 1984; Dods, 1987) and instrumented vehicles (Allen, Hogue, Rosenthal, & Parseghian, 1988; Blaauw & Burry, 1980). In the first case the driver is unaware that he or she is participating in an experiment, whereas in the second case it could be arg ued that because the driver must be aware of the experiment, it is probable that his or her behavior is affected. As far as videocameras are concerned, placement and calibration of the cameras require special attention if high measurement accuracy is needed.
When a simulator is driven, monitoring and measurement of vehicle lateral position are recorded automatically and no accuracy problem exists. Therefore, investigating the vehicle's path in the simulator is an easy task for a researcher. Driving simulators have been widely used by traffic psychologists to investigate human factors issues such as the effects of drugs (Brookhuis, Volkerts, & O'Hanlon, 1990; Gawron & Ranney, 1990; O'Hanlon, Haak, Blaauw, & Riemersma, 1982) and fatigue (Desmond & Matthews, 1996) on driving performance. Using the standard deviation of lateral position to measure a driver's ability to control the weaving of the car appears to be a very sensitive indicator of drug-induced sedation.
Lateral position on an actual road has been investigated mainly in the context of road geometry. For example, Glennon and Weaver (1971) and Glennon, Neuman, and Leisch (1985) investigated vehicle path in terms of centerline encroachments; McLean (1974) and Johnston (1983) in terms of corner-cutting strategies; Neuhardt, Herrin, and Rockwell (1971) in relation to road curvature, lane width, and curve length; and Reinfurt, Zegeer, Shelton, and Neuman (1991) with regard to edge-line encroachments.
Although the foregoing review reveals considerable research on drivers' lateral position on the road and in the simulator, there are few studies (Blaauw, 1982; Blana, 2001; Harms, Alm, & Tornos, 1996; Reed & Green, 1995) comparing drivers' lateral position in the two environments under the same circumstances (e.g. type of road, road geometry and environment, traffic conditions; Blaauw, 1982; Blana, 2001; Harms, Alm, & Tornos, 1996; Reed & Green, 1995).
This article attempts to define the relation between lateral position on the actual road and in a fixed-base driving simulator and to investigate the cause of any differences that may be found. It also attempts to indicate the benefits of knowing this relation when conducting experiments in the driving simulator.
The simulator participants were 100 drivers selected from the general driving population of Leeds, England, to represent people of either gender and all reasonable ranges of age and driving experience. The final sample consisted of 50 men and 50 women with an average age of 36 years and driving experience of at least 3 years. All participants were given a practice run of 15 mm to become accustomed to the simulator controls.
For the real-road study, 100 drivers (50 men, 50 women) of free-flowing, medium-size passenger vehicles were observed. …