A novice driver transfer of training experiment

This paper gives an overview of a five year transfer of training study of novice drivers in California, USA. The study included simulator training, analysis of training data and a longitudinal analysis of participant crash history as they obtained their license and drove in the real world. Crash data was obtained from the California Department of Motor Vehicles for a three year period. Three groups of novice drivers were trained on three simulator configurations with differing levels of fidelity. Crash rates are compared for these three groups with previously published crash rate data obtained for traditionally trained novice drivers. The results show that simulator fidelity has significant effect on training effectiveness and that it is possible to reduce the high novice driver crash rate with simulator training. This paper explores these results from a transfer of training and simulator fidelity and validity point of view. Directions for future training studies are also provided.