Traffic Accident Data, New South Wales, 1991

This study was conducted to develop a statistical model of the prediction of the severity and type of motor vehicle accident injuries. This model could then be examined to determine which variables were most influential in the severity of injuries and these areas could be targetted by police to reduce the occurrence of serious injury and death. The data used consisted of observations of injuries sustained on NSW roads throughout 1991. It takes into account their relationship with several key factors, a few of which are generally well known, but several that are not. Main variables noted by police at the scene of the accident were such things as the use of seatbelts, vehicle make, type and age, speed of vehicle, seating position of caualty, blood alcohol reading, type of collision and time of day of accident. Background variables include age and sex of the victim.

本研究旨在构建机动车事故伤害严重程度与伤害类型的预测统计模型。后续可通过对该模型的分析，明确对伤害严重程度影响最为显著的变量，警方可针对这些关键领域开展针对性干预，以降低严重伤害与死亡事件的发生风险。本研究所用数据集涵盖1991年全年新南威尔士州（New South Wales, NSW）道路上的人员伤害观测记录，该数据集考量了伤害与多项关键影响因素之间的关联，其中部分因素已被广泛认知，另有部分因素则尚未得到充分关注。警方在事故现场记录的核心变量包括：安全带使用情况、车辆品牌、车型与车龄、车辆行驶速度、伤者座位位置、血液酒精含量、碰撞类型以及事故发生时段。背景变量则涵盖受害者的年龄与性别。