Transfer of Skills Learned on a Driving Simulator to On-Road Driving Behavior

Abstract

A long-term, naturalistic, prospective-cohort transfer of training study was conducted in commercial driving schools in Quebec, Canada, to test the effects on driver performance and behavior of integrating driving simulator–based training (DSBT) into the driver education program. For the study, 1 h of DSBT could be substituted for 1 h of on-road training for up to six of the mandatory 15 h. Four driving schools provided a convenience sample of 1,120 learner drivers (average age 17.7 and 52.7% female) between January 2010 and December 2014. Of the study sample, 95% received 1 to 4 h of DSBT. Those in the comparison group were all new, young Quebec drivers who had completed the mandatory driver education program in the same period. This paper reports on the association between DSBT and government driving records for 2 years after licensing. The DSBT group recorded lower infraction rates and, controlling for vehicle ownership and age, comparable crash rates. The lower infraction rates for males, despite the higher vehicle ownership normally associated with greater and riskier driving exposure, are a positive and unexpected finding. Crashes are multifactorial events less obviously related to drivers’ skills or intentions, and the comparable crash rates potentially indicate absence of overconfidence attributable to a form of advanced driver training. Overall, these results show that the substitution of relatively few hours of DSBT for on-road training is associated with reduced infractions and has no apparent influence on crashes in the first 2 years of unsupervised driving after licensing.