Assessing distracted driving crash severities at New York City urban roads: A temporal analysis using random parameters logit model

Sina Rejali*, Kayvan Aghabayk, Mohammad Ali Seyfi, Oscar Oviedo-Trespalacios

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

1 Citation (Scopus)
17 Downloads (Pure)

Abstract

Distracted driving poses one of the most significant risks to road safety. The current study aims to provide a deeper understanding of the factors affecting the severity of distracted driving crashes in New York City and to explore the temporal stability in the effects of different variables on crash outcomes in 2016 to 2019 period by adopting a post-crash perspective. The police-reported data of single-vehicle distraction-related crashes of private cars on urban roads of New York City was used for this study. Three injury categories were considered: no injury, minor injury, and severe injury. To investigate crash severities and identify unobserved heterogeneities, a random parameters logit model was conducted. The results revealed that a wide variety of variables including driver traits, vehicle and temporal characteristics, and crash attributes were found to be significant in explaining distracted-related crash severities. Furthermore, a series of likelihood ratio tests were conducted to identify the temporal shifts of estimated variables during the period. The results of the temporal analysis showed that the estimated variables of the random parameters model were unstable during the 4-year period, which may be the result of shifting trends such as the development of in-vehicle technologies, and new sources of distraction. However, the complex nature of distracted-related crashes and changes in driver behavior should be considered for further interpretation. This research provides a set of policy implications for planners and policymakers, aiming at facing factors contributing to a higher level of injury severity in distracted driving crashes. This includes providing targeted information on distracted driving to high-risk groups, such as younger drivers, and also combining education, awareness programs, higher penalties, and intense patrolling. Engineering measures such as enhanced roadside illumination and audible edge lines can be effective, especially in reducing late-night distracted driving crashes.

Original languageEnglish
Pages (from-to)147-157
Number of pages11
JournalIATSS Research
Volume48
Issue number2
DOIs
Publication statusPublished - 2024

Keywords

  • Crash severity
  • Distracted driving
  • Multitasking
  • Random parameters logit
  • Temporal assessment

Fingerprint

Dive into the research topics of 'Assessing distracted driving crash severities at New York City urban roads: A temporal analysis using random parameters logit model'. Together they form a unique fingerprint.

Cite this