THE VALIDITY OF SIMULATORS IN STUDYING DRIVING BEHAVIOURS
Road crashes are still the major cause of traumatic death and injury in Australia with an economic and social burden estimated at over $18 billion in 2009. Along with other approaches to road safety, policymakers and authorities in developed countries have turned to Intelligent Transport Systems’ infrastructure such as advanced driving simulators to investigate road safety issues. Driving simulators can be a relatively low-cost way of researching human behaviour and road safety issues without the logistical demands of on-road research. As a consequence they have the potential to explain interactions between drivers and other drivers, vehicle features and roadway surroundings, and to explore effective countermeasures to reduce traffic crash rates. Simulators range from very simple apparatus through to quite sophisticated equipment which recreates the driving task as accurately as possible with a correspondingly wide range of costs and features. However, the key question about the usefulness of simulator research is how valid the results are, i.e. the degree to which the research results obtained in simulators reflect on-road driving behaviour. This paper presents a critical review of studies which have compared the results of simulator research with on road behaviours including, speed, driving while fatigued, driving while distracted, drink driving and drug driving. The findings are discussed in relation to issues such as the nature of the behaviour studied, the physical fidelity of the simulators and characteristics of the simulation (e.g. geometric design) and makes recommendations for the value of future simulator-based research.