Gender Differences in Injury Severity Risk of Single-Vehicle Crashes in Virginia: A Nested Logit Analysis of Heterogeneity

Abstract: This paper explores gender differences in injury severity risk using a comprehensive crash dataset including the driver, vehicle, environment, and roadway characteristics. For the purpose of this study, only single vehicle crashes that occurred in the Commonwealth of Virginia, collected in a five year period from 2011 to 2015, were used. This study contributes to the literature of crash analysis by incorporating an extensive dataset with normalized attributes. The dataset used for model estimation integrated from two different data sources. These data sources include Virginia Traffic Records Electronic Data System (TREDS) to incorporate crash information and Virginia Base Mapping Program to incorporate roadway characteristics. A two level nested logit model is developed for each gender, in order to relax the Independence of Irrelevant Alternatives (IIA) assumption. Several crash determinants including but not limited to, drivers characteristics, drivers behavior, vehicle condition, weather conditions, roadway geometry and surface conditions were found to have an impact on both male and female drivers. This study is an assessment of influence of different factors on crash severity segmented by gender. A few interesting results include; (i) female fatality risk exceeds male fatality risk when driving under the influence of alcohol, (ii) speeding contributes to lower fatality risk among female drivers compared to their male counterparts, (iii) poor vehicle conditions have no impact on injury risk to female drivers while having an increased injury risk among male drivers, and (iv) while driving in a work zone area increases the risk of property damage crashes, the impact is higher for male drivers compared to female drivers.