Shechtman Orit, Classen Sherrilene, Awadzi Kezia, Mann William
Department of Occupational Therapy and National Older Driver Research and Training Center, College of Public Health and Health Professions, University of Florida, Gainesville, Florida 32611, USA.
Traffic Inj Prev. 2009 Aug;10(4):379-85. doi: 10.1080/15389580902894989.
Driving simulation provides a convenient and safe method for assessing driving behaviors. Many authors, however, agree that validation is a key component of any study that utilizes simulators to assess driving performance. The purpose of this study was to test driver response validity by discerning whether behavioral responses of drivers, as expressed by type and number of errors, are similar on the road and in the simulator.
We replicated real-world intersections in our driving simulator (STISIM M500W; Systems Technology Inc.) and assessed the number and type of driving errors committed by the same 39 participants while negotiating a right and a left turn both on the road and in the simulator.
We found no significant interactions between the type of vehicle (road vs. simulator) and the type of turn (right versus left) for any of the driving errors, indicating that the same trends exist between driving errors made on the road and in the simulator and thus suggesting relative validity of the simulator. We also found no significant differences between the road and the simulator for lane maintenance, adjustment to stimuli, and visual scanning errors, indicating absolute validity for these types of errors.
The findings suggest early support for external validity for our driving simulator, indicating that the results of assessing driving errors when negotiating turns in the simulator can be generalized or transferred to the road under the same testing conditions. A follow-up study with larger sample size is needed to establish whether driving performance in the simulator is predictive of driving performance on the road.
驾驶模拟为评估驾驶行为提供了一种便捷且安全的方法。然而，许多作者一致认为，验证是任何利用模拟器评估驾驶性能的研究的关键组成部分。本研究的目的是通过辨别驾驶员的行为反应（以错误类型和数量表示）在道路和模拟器中是否相似，来测试驾驶员反应的有效性。
我们在驾驶模拟器（STISIM M500W；系统技术公司）中复制了现实世界中的十字路口，并评估了39名相同参与者在道路和模拟器中进行右转弯和左转弯时所犯驾驶错误的数量和类型。
对于任何驾驶错误，我们发现车辆类型（道路与模拟器）和转弯类型（右转弯与左转弯）之间均无显著交互作用，这表明在道路和模拟器上所犯的驾驶错误之间存在相同的趋势，从而表明模拟器具有相对有效性。我们还发现，在车道保持、对刺激的调整以及视觉扫描错误方面，道路和模拟器之间没有显著差异，这表明这些类型的错误具有绝对有效性。
研究结果为我们的驾驶模拟器的外部有效性提供了早期支持，表明在相同测试条件下，在模拟器中转弯时评估驾驶错误的结果可以推广或转移到道路上。需要进行一项样本量更大的后续研究，以确定模拟器中的驾驶性能是否可预测道路上的驾驶性能。
