Application of Multinomial Regression Model to Identify Parameters Impacting Traffic Barrier Crash Severity

Rezapour, Mahdi; Molan, Amirarsalan Mehrara; Ksaibati, Khaled · 2019 · Crossref

DOI: 10.2174/1874447801913010057

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the factors influencing the severity of traffic barrier crashes, specifically focusing on Run-Off-The-Road (ROTR) incidents in Wyoming. While traffic barriers are critical countermeasures for roadside safety, crashes involving them still result in significant fatalities and severe injuries. The research was motivated by the need to identify unique contributory factors for barrier crashes, which differ from other crash types, and to address inconsistencies in previous literature regarding variables like speed limits. The study aimed to provide policymakers with evidence-based directions for mitigating crash severity. The researchers utilized a multinomial logistic regression model to analyze 10 years of crash data (2007–2016) obtained from the Wyoming Department of Transportation. The dataset included crashes where vehicles struck traffic barriers as the first harmful event. To account for operational differences, the analysis was split into two distinct highway classes: interstate and non-interstate systems. Crash severity was categorized into three levels: severe (fatal or incapacitating injury), minor (possible or non-incapacitating injury), and property damage only (PDO), which served as the reference category. The proportional odds assumption was tested and rejected, confirming the appropriateness of the multinomial model over an ordered logistic regression. Variables examined included driver characteristics (age, gender, restraint use, citation records), environmental conditions (road surface, weather), and roadway features (barrier type, curve negotiation). For non-interstate highways, four factors were statistically significant. Adverse road conditions (wet, snow, slush, ice) significantly decreased the odds of severe crashes compared to PDO crashes, likely due to more cautious driving behavior. However, older drivers (≥35 years), improper driver restraint, and negotiating horizontal curves significantly increased the odds of severe crashes. Specifically, improper restraint increased the odds of severe injury by approximately 4.2 times, and negotiating a curve increased the odds by 3.4 times. For interstate highways, the type of barrier was a significant factor; cable barriers reduced the odds of severe crashes compared to other barrier types, while concrete barriers increased the odds of minor crashes. Additionally, driver condition (non-normal), prior citation records, and speed limit compliance were identified as significant predictors of crash severity on interstates. The findings highlight that crash severity determinants vary significantly between interstate and non-interstate environments. The study concludes that targeted countermeasures, such as improving barrier conditions on curves and enforcing restraint use, can alleviate severity on non-interstate roads. On interstates, the superior performance of cable barriers suggests potential benefits in upgrading barrier types. These results offer specific guidance for transportation agencies to enhance roadside safety designs and policies.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• crash typology
• pre crash contributing factors

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: crash risk outcomes