ROAD ACCIDENT SEVERITY ANALYSIS - A MICROLEVEL APPROACH

A comprehensive understanding of the relationship between road acciden t occurrence and severity of consequences permits the formulation of s afety measures that are most cost-effective. A disaggregate model of r oad accident severity based on sequential logit models is presented. T he sequential binary approach is able to account for the dependency be tween different levels of severity. Factors that affect the level of d amage experienced by individuals involved in road accidents include th e following: accident dynamics, seating position, vehicle condition, v ehicle size, driver condition, and driver action. Separate models are calibrated for three accident situations: single-vehicle accidents, tw o-vehicle accidents, and multi-vehicle accidents. Ontario road acciden t police reports are used to calibrate and validate the models. The re sults of a simple application of the models to a safety protocol invol ving the effectiveness of passenger restraint devices are presented.