One of the many important tasks facing traffic safety managers is deciding upon which, and of these the ‘best/optimum’ measures/countermeasures to implement for addressing their main goal -- improving road safety. The ability to make these decisions is largely dependent upon the availability of relevant, accessible, timely and standardized data on the incidence of travel and occurrence of vehicle collisions on the roads and highways. In essence, the Road Safety Directorate of the Canadian federal Department of Transport is continually striving to understand and augment its knowledge with respect to the process of motor vehicle collision causation in order to recognize opportunities for avoiding accidents and reducing casualties. An area crucial to traffic safety research is the development of evaluation methods for measuring and subsequently identifying ‘high risk’ road user groups and their associated travel patterns and characteristics. This component of the countermeasure development process is difficult to pursue since the overall road travel risk of accident occurrence is directly affected by the joint interactive and ever-changing effects of numerous driver-passenger-vehicle-road-tripenvironment- temporal travel pattern risk levels existing within the transportation system. Other factors including implemented countermeasure programs, economic conditions, vehicle/driver regulations, and social factors also influence the prevailing risk levels, at any given time, for road systems users.

It is useful to define the risk levels associated with all of the various factors contributing to accidents by comparing their appearance in accidents with some comparable measure of their appearance in traffic. This latter measure we refer to as the ‘exposure to risk’ associated with the factors, i.e., the extent to which they are exposed to the possibility of accidents, by appearing in traffic. Although a variety of exposure measures have been advocated (e.g., trip frequencies, driver/vehicle frequencies, travel time, etc.) the most suitable for describing and comparing road users ’ exposure is driver and passenger kilometers of travel -- the exposure measure that is recommended and used in this study.

This paper presents the results for five main objectives identified in a recent road travel risk research study conducted by Transport Canada. The ‘first objective involved the development and implementation of a statistical methodological framework that combines collision and ‘exposure to risk’ data to measure and interpret risk performance indicators. Secondly, a ‘risk analysis and evaluation system modeI’ for evaluating, comparing, and monitoring the relative risks of collision, injury and fatality encounter associated with the various driver-passenger-vehicie-road-trip-environment-temporal travel patterns and characteristics was developed. The third objective involved the development and implementation of statistical methods and procedures for measuring levels of errors associated with the various types of road travel risk, reIative risk, and relative risk odds-ratio perfortnance indicator estimators to identify significant differences in travel risks prevailing on the roads and highways. Fourthly, using Canadian collision and ‘exposure to risk’ data bases, the modeling framework was applied to measure the relative risks of collision, injury and fatality encounter for selected road user groups and their travel patterns and characteristics. From these results ‘high risk’ road travel patterns and characteristics were identified. Finally, conclusions and recommendations regarding the.uses and applications of the ‘risk analysis and evaluation system model’ for identifying road travel problem areas/issues and evaluating remedial measures/countermeasures for improving road safety are provided.