R.W. Allen, G.D. Park, M.L. Cook, D. Fiorentino

Pages: 51-68

Abstract
Purpose: To determine whether driving simulator training and simulator fidelity can influence novice driver crash rates. Methods: Over 500 teenaged novice drivers in a between-groups experimental design were trained on three different simulator configurations with varying degrees of fidelity: 1) a single-monitor desktop, 2) threemonitor desktop, and 3) an instrumented vehicle cab. The simulator driving scenarios were set up to give repeated exposure to critical situations involving traffic, pedestrians and traffic control devices. Subject selection did not involve random assignment: training with the first simulator configuration was conducted at high school sites and driver education students were required to take the training; participants training with the second and third simulator configurations were volunteers recruited at California DMV offices and the training occurred in research laboratories. Results: Simulator training procedures and results are reported along with California Department of Motor Vehicle crash data obtained roughly two years after the end of training. Results showed that participants trained on higher fidelity simulator configurations had significantly lower crash rates than conventionally trained novice drivers as reported in the literature. Discussion: Discussion includes the effects of simulator fidelity, driver characteristics, training schedule and environment as possible influences on simulator training efficacy. Limitations of the experimental design and procedures are discussed as a guide for future work in this area.

Keywords: novice driver training; simulation; computer based instruction