Analyzing Crash Severity near Freeway Entrance Ramps in North Carolina Using a Hierarchical Ordered Probit Model

Freeway entrance ramps have long been considered crash-prone locations. Understanding factors influencing the severity of crashes near entrance ramps can help improve freeway safety. This study analyzed crash severity near entrance ramps using a hierarchical ordered probit (HOPIT) model, which accounts for threshold heterogeneity, offering a more flexible and precise assessment of roadway, environmental, and driver-related factors. Crash data from North Carolina’s Highway Safety Information System (HSIS) from 2013 to 2017 were analyzed, covering 2,974 crashes. Key variables included driver age, gender, crash type, roadway design, environmental conditions, and trafficway type. The study also conducted a comparative assessment of the HOPIT, ordered logit, and ordered probit models, and confirmed the statistical superiority of the HOPIT model in capturing variability in severity thresholds. The findings indicated that sideswipe, rear-end, and fixed-object crashes significantly increased crash severity. Male drivers were more likely to be involved in severe crashes than were female drivers. Undivided trafficways and truck involvement also contributed to higher crash severity. Weekend crashes significantly impacted the likelihood of transitioning from minor injuries to severe injuries, highlighting the role of behavioral and temporal factors. Conversely, variables such as weather conditions, speed limits, shoulder width, and annual average daily traffic (AADT) did not have significant effects, suggesting that driver adaptation and real-time roadway conditions played a more influential role. These insights contribute to the academic discourse on traffic safety and hold practical implications for traffic management and infrastructure design. The results underscore the need for targeted safety interventions and infrastructure improvements to enhance freeway safety at entrance ramps.