USING VISUAL SIMULATION TO EVALUATE BICYCLISTS' PERCEPTIONS OF SELECTED RISK FACTORS

Casual and experienced bicyclists were immersed in a 'virtual' or computer-generated simulation of a two-lane roadway environment in order to elicit ratings of the perceived risk associated with various lane conditions as well as different vehicle speeds and volumes. Ratings were made under cyclist, driver, and roadside viewing conditions. Levels of perceived risk varied inversely with lane width as well as the presence/absence of lane markings delineating vehicle from non-vehicle areas of operation. Effects were most pronounced for older subjects (over 20) and under conditions where ratings were made from a driver's eye point. With respect to the influence of vehicle speeds and volumes on perceived risk, speed exerted the more pronounced effect. Outcomes obtained in the present study under virtual conditions showed the same functional relationships between risk and design variables as those obtained under actual roadway conditions (Harkey and Knuiman, 1997), although there was a trend toward lower levels of perceived risk under virtual conditions. The results suggest that simulation may, where appropriate, represent a cost effective alternative to the use of traditional field data collection methods, with the chief advantages being the degree of experimental control possible over dynamic variables and the ability to systematically vary the presence/absence of key design variables.