About 30% of casualty accidents in the Melbourne metropolitan area occur at arterial/local intersections, which is more than occur at arterial mid-block locations, at arterial/arterial intersections, or on the entire local street system. Since little was known about these accidents, a screen-based procedure for interrogating police accident files containing records of interview and statements as well as accident report forms was developed. The procedure asked about very specific aspects of accident participants' behaviour, including the manner and speed with which they approached the accident situation, whether there were distractions, whether they saw the other vehicle, whether they had looked, whether they made wrong assumptions about what the other traffic unit would do, and what they did to avoid the collision. Files relating to seven accident types based on traffic movements were examined. Not seeing the other traffic unit in time to avoid a collision was the most frequent failure identified. The percentage of drivers not seeing the other traffic unit varied with accident type. At least one driver failed to see the other vehicle in 69-80% of vehicle-vehicle collisions, and 33% of drivers in single vehicle accidents failed to detect the hazard collided with. There was little attempt to attribute faulty behaviour such as speeding or failure to signal to the other driver. In many cases, the driver looked in the appropriate direction but failed to see the other vehicle. In other cases, the view of the approaching vehicle was obscured by other traffic. The implications of these findings for countermeasures are discussed, including kerb extensions and median treatments, daytime running lights, and driver education. (A)
Abstract