Purpose of this study was to determine the validity of a part task driving simulator. As a preliminary step, many validity concepts applicable to simulation devices were explored both analytically and experimentally. Validation of the driving simulator was then attempted in terms of: 1. Correspondence between selected real-world and simulator driving behaviours: 2. statements obtained from subjects concerning the face validity of the device; and 3. an engineering evaluation of simulator display and control fidelity. Behavioural measures for a variety of comparable driving conditions were collected both in simulated and real-world environments. Overtaking and passing performance was studied under 1. three levels of ambient illumination, 2. high and low rates of overtake, and 3. three different taillight configurations at night. Vehicle following behaviour was also measured under day and night driving conditions. Results indicated that empirically derived performance estimates of validity may be employed for simulator validation. In the present similar a high degree of absolute correspondence was not obtained between simulator and real-world performance measures. However a reasonably high degree of relative correspondence was obtained- The latter correspondence may be high enough to permit acceptable predictions of real-world performance from the corresponding performances measured in the simulator. Estimates of face validity wore also high. The simulator appeared to duplicate a typical vehicle in a rural, two-lane highway driving environment. Results of an engineering evaluation of simulation fidelity confirmed this impression- The display and control fidelity of the device is generally high.
Samenvatting