A binary logistic regression model of the driver avoidance manoeuvers in two passenger vehicle crashes

Hayder Mohammed Hayder Al-Taweel, William Young, Amir Sobhani

Research output: Contribution to conferencePaper

4 Citations (Scopus)

Abstract

The reactions of drivers may influences the risk and severity of car crashes. However, while studies have analysed drivers’ reactions in crashes, in general, little is known about the factors affecting crash avoidance maneuvers of two passenger vehicle crashes. To increase understanding in this area a statistical model is developed using the National Automotive Sampling System-Crashworthiness Data System (NASS-CDS) between the years 2009 and 2014. More specifically, a Pearson Chi-Square test is performed first to identify the significant variables, and then a Binary Logistic Regression Model is developed to identify the relative importance of the variables. The results of the model indicate that, elderly drivers or (65 years or older), are less likely to avoid crashes than other groups. Drunk or drugged drivers have a greater negative effect on engaging in crash avoidance maneuvers. Drivers of large size vehicles are more likely to react than those in smaller cars. Driving in adverse surface conditions, non-level profile, and rural roads, increases the likelihood of reacting. Hitting drivers have higher possibility to react than those in the hit vehicle. In future research, it is encouraged to investigation the relation between the driver’s reaction for hit and hitting drivers and crash severity using Binary Logistic Regression Model.

Original languageEnglish
Publication statusPublished - 2016
EventAustralasian Transport Research Forum 2016 - Melbourne, Australia
Duration: 16 Nov 201618 Nov 2016
Conference number: 38th
https://www.australasiantransportresearchforum.org.au/papers/2016 (Proceedings)

Conference

ConferenceAustralasian Transport Research Forum 2016
Abbreviated titleATRF 2016
CountryAustralia
CityMelbourne
Period16/11/1618/11/16
Internet address

Keywords

  • Binary Logistic Regression Model (BLRM)
  • Crash avoidance maneuvers

Cite this