Previous studies have commonly identified a significant difference in injury severity in traffic crashes by gender. In the United States, using a female dummy had not been mandatory in vehicle testing until 2003 for frontal crashes and 2006 for side impact crashes. The main objective of this study is to identify the gender difference in injury severity by considering the use of female dummy in vehicle safety testing.

Two random parameter logistic models are developed using crash data of Florida. The results demonstrate that female drivers in the vehicle made after 2003 are less likely to be seriously injured, whereas male drivers in the vehicle made after 2006 have a lower propensity to be seriously injured. More importantly, gender disparities in injury severity still exist and are even larger when considering a female dummy in testing. Other variables to control the effects are found significant including age, alcohol/drugs use, seat-belt use, vehicle type, crash type, road features and other environments. In conclusion, this study supports that newer vehicle models in this study are safer for males and females but still there is a room for reducing gender disparities in traffic safety.

先前的研究通常按性别确定交通事故中伤害严重程度的显着差异。在美国，直到2003年正面碰撞事故和2006年侧面碰撞事故之前，在车辆测试中都没有强制使用女性假人。这项研究的主要目的是通过考虑在车辆安全性测试中使用女性假人来识别伤害严重程度的性别差异。

使用佛罗里达州的崩溃数据开发了两个随机参数逻辑模型。结果表明，2003年以后生产的车辆中的女性驾驶员受重伤的可能性较小，而2006年以后生产的车辆中的男性驾驶员受重伤的可能性较低。更重要的是，在测试中考虑女性假人时，伤害严重程度方面的性别差异仍然存在，甚至更大。发现控制效果的其他变量也很重要，包括年龄，酒精/毒品使用，安全带使用，车辆类型，碰撞类型，道路特征和其他环境。总之，本研究支持本研究中较新的车辆模型对男性和女性更安全，但在减少交通安全中的性别差异方面仍有空间。