Citation

Chen Z, Lym Y. Transp. Policy 2021; 101: 34-45.

Copyright

(Copyright © 2021, Elsevier Publishing)

DOI

10.1016/j.tranpol.2020.11.011

PMID

unavailable

Abstract

This study evaluates the influences of built environment on frequency and severity of vehicle crashes through a comparative analysis framework between crashes involved in distracted driving (DD) and non-distracted driving (Non-DD). Using a comprehensive dataset with 1.4 million vehicle crash records in Ohio for the period 2013-2017 as an example, the relationship between built environment and the frequency as well as severity of vehicle crashes caused by DD and Non-DDwas examined using negative binomial and generalized ordered logit regression methods. Our study reveals that the frequency of DD crashes tends to be much higher than that of Non-DD crashes in urbanized road environment. Conversely, other road features, such as a median and a shoulder with an asphalt pavement are found to have relatively substantial and negative associations with the frequency of DD related crashes. Furthermore, the outcomes of severity analysis confirm that DD related crashes tend to be more severe than Non-DD related crashes in certain built environment such as work zones or interstate highways. Meanwhile, the existence of a roundabout was found to have mixed effects in affecting crash severities. Our findings suggest that distraction affected crashes contain unique features as compared to Non-DD counterparts. These evidences provide important implications for transportation planners and decision-makers to improve roadway safety.

Language: en

Keywords

Built environment in Ohio; Crash frequency and severity; Distracted driving; Generalized ordered logit model; Negative binomial model