Reprint of "Modeling two-vehicle crash severity by a bivariate generalized ordered probit approach"

Area of Science:

• Traffic Safety
• Transportation Engineering
• Econometrics

Background:

• Signalized intersections are critical points for traffic accidents.
• Understanding factors influencing crash severity is crucial for developing effective safety measures.
• Previous models often analyze crash severity unidirectionally, potentially missing complex interactions.

Purpose of the Study:

• To simultaneously model the crash severity of both parties in two-vehicle accidents at signalized intersections.
• To introduce and validate a novel bivariate generalized ordered probit (BGOP) model for crash severity analysis.
• To identify key risk factors contributing to different crash severity levels.

Main Methods:

• Development and application of a bivariate generalized ordered probit (BGOP) model.
• Simultaneous modeling of crash severity for both involved parties in two-vehicle collisions.
• Analysis of crash data from signalized intersections in Taipei City, Taiwan.

Main Results:

• The BGOP model demonstrated superior goodness-of-fit and prediction accuracy compared to the conventional bivariate ordered probit (BOP) model.
• Key risk factors identified include older drivers (age > 65), motorcycles, alcohol use violations, specific intersection geometries (three-leg, multiple-leg), rear-ended collisions, and nighttime lighting conditions.
• The model effectively identified factors influencing varying crash severity levels.

Conclusions:

• The BGOP model offers a more robust approach for analyzing crash severity in two-vehicle accidents.
• Specific driver, vehicle, violation, intersection, and environmental factors significantly increase crash severity.
• The findings provide a basis for targeted safety interventions and policy development to reduce traffic accidents at signalized intersections.