E.K. Adanu, S.L. Jones, A.D. Lidbe

Pages: 89-100

Abstract
In this study, the effects of spatial, temporal, and human-centered factors on alcohol-impaired crash outcomes are explored. The study was based on 2016-2017 Alabama crash data obtained from the Critical Analysis Reporting Environment system developed by the Center for Advanced Public Safety at the University of Alabama. To capture the effects of unobserved heterogeneity due to randomness associated with factors necessary to understand alcohol-impaired crash injury severity, mixed logit modeling technique was used. Location, time of crash, and driver attributes have been found to affect injury severity of crashes. The model estimation results show that alcohol-impaired crashes that occurred between 6PM and midnight in rural areas were more likely to result in severe injuries but those that occurred on Fridays for the same period were less likely to result in injuries. Urban locations were found to be less likely to record severe injury crashes during weekends and summer and minor injury crashes were more likely to occur between midnight and 6AM on weekends. At-fault drivers who engaged in alcohol-impaired driving with no valid license and failed to use seatbelts had higher chances to be severely injured. The study identified high risk regions and characteristics of at-risk drivers for effective and efficient implementation of strategies to reduce alcohol-impaired crashes. Ultimately, this study contributes to efforts in understanding and changing risky road user behaviors as critical components of the holistic safe systems approach to achieving road safety goals.
Keywords: alcohol-impaired crash; human-centered; spatio-temporal; crash severity; random parameters logit