experimental design reflecting subject attributes and trial randomization. A major improvement would be the development and application of multivariate statistical models that allow the study of interrelations between variables.
An example of such a model is the use of predicted speed as an instrumented variable in reaction time models. This formulation captures the notion that speed is endogenous to reaction time--that is, it is an outcome of a driver decision-making process, as is reaction time. The proper model would not treat speed as an independent variable, but as a variable chosen by the driver and conditions the reaction time observed. Although drivers are asked to drive close to the speed limit, the relative fidelity of the simulation may lead them to choose speeds affecting reaction times. A model of the type described would more precisely control for that effect.
Research needs to be conducted using older subject populations. This study used relatively young subjects (between ages 30 and 40). Further research is needed with alternative formats of turn-by-turn displays for both head-up and head-down modes. This would help in determining whether a head-up display leads to better driving performance compared to a head- down display, given a particular display format. Research is also needed to explore whether a simpler electronic map leads to better driving performance compared to a complex electronic map. Finally, experiments need to be conducted on-the-road to validate results from simulator experiments.
The authors would like to thank t he Office of Competitive Technology of the California Department of Commerce for their financial support during this study. The authors are indebted to Cheryl Hein, Barry Berson, Francine Landau, Mike Laur, Craig Lee, and many other employees of the Hughes Aircraft Company whose contribution was invaluable in the development of the driving scenes and data collection protocols.
Burnett G. E., & Joyner S. M. ( 1993). "An investigation of man machine interfaces to existing route guidance systems". Proceedings of the 4th International Conference on Vehicle Navigation and Information System, 395-400. Piscataway, NJ: IEEE.
Hart S. G., & Staveland L. E. ( 1988). "Development of NASA-TLX (Task Load Index): Results and theoretical research". In P. A. Hancock & N. Meshkati (Eds.), Human mental workload (pp. 1 39)-1 83)). Amsterdam: Elsevier Science.