Exploring the overall and specific crash severity levels at signalized intersections.

Area of Science:

• Road safety
• Traffic engineering
• Transportation research

Background:

• Intersections are high-risk locations in road networks.
• Understanding factors influencing crash injury severity is vital for developing effective safety measures.
• Crash data quality and completeness can significantly impact analysis outcomes.

Purpose of the Study:

• To identify factors affecting crash injury severity at signalized intersections.
• To evaluate the impact of data completeness on injury severity modeling.
• To compare different modeling techniques for crash injury analysis.

Main Methods:

• Utilized ordered probit modeling to analyze naturally ordered injury levels.
• Employed tree-based regression to explore factors influencing each severity level.
• Cross-checked data from multiple sources to ensure completeness, including minor crashes.

Main Results:

• Crash-specific information and intersection characteristics significantly predict injury level.
• Pedestrian/bicyclist involvement and left-turn crashes are associated with higher injury severity.
• Divided minor roadways and higher speed limits on minor roads correlate with reduced injury severity.
• Data completion with minor, non-injury crashes improved modeling accuracy.

Conclusions:

• Signalized intersection safety can be enhanced by addressing factors like pedestrian/bicyclist presence and left-turn maneuvers.
• Road design elements such as divided minor roadways and appropriate speed limits can mitigate injury severity.
• Accurate and complete crash data are essential for reliable injury severity modeling and effective safety interventions.