Influential Factors Associated with Consecutive Crash Severity: A Two-Level Logistic Modeling Approach.

Area of Science:

• Traffic Safety
• Transportation Engineering
• Accident Analysis

Background:

• Consecutive crash series, involving primary and secondary impacts, differ from typical chain-reaction or queuing crashes.
• These events impact large road areas, leading to congestion and clearance delays.
• Identifying factors influencing consecutive crash severity is crucial for road safety.

Purpose of the Study:

• To identify influential factors determining the severity of primary and secondary crashes within a consecutive crash series.
• To compare the performance of different logistic regression models in analyzing consecutive crash data.
• To provide policy recommendations for mitigating consecutive crash severity.

Main Methods:

• Collected crash data from the freeway network of Guizhou Province, China in 2018.
• Identified 349 consecutive crash series from the dataset.
• Established and compared basic, random-effects, random-parameters, and two-level binary logistic regression models.

Main Results:

• The two-level logistic model demonstrated superior performance in analyzing consecutive crash data.
• At the crash level, double-vehicle primary crashes negatively associated with secondary crash severity; truck involvement positively contributed to secondary crash severity.
• Road segment factors like speed limit, traffic volume, tunnels, and adverse weather (rain, clouds) increased severity, while more lanes decreased it.

Conclusions:

• The two-level logistic model effectively identifies key factors influencing consecutive crash severity.
• Specific crash-level (e.g., truck involvement) and road-level (e.g., weather, speed limits) factors significantly impact crash severity.
• Targeted interventions, including real-time hazard warnings and driver education, can help reduce the severity of consecutive crashes.