ASSESSMENT OF HIGHWAY GEOMETRIC CHARACTERISTICS ON TRUCK CRASH SEVERITY ALONG SELECTED NIGERIAN HIGHWAYS USING MIXED LOGIT MODELS

Summary

This empirical research investigated the contributory interactions among highway geometric design parameters, operational traffic characteristics, and environmental factors on heavy freight vehicle crash severity along critical Nigerian transport corridors. The methodological framework integrated primary data acquisition via 390 validated stakeholder questionnaires, a GIS-based spatial kernel density estimation, and discrete choice econometric modeling. Spanning the period from January 2018 to December 2025, an initial population of 3,850 historical crash records was retrieved from the Federal Road Safety Corps (FRSC) National Crash Reporting Information System (NACRIS) database for the selected study corridors (the Lagos – Ibadan Expressway and Benin – Ore Highway). Following a rigorous data filtration process comprising the elimination of duplicate entries, incomplete geometric logs, and missing severity descriptors (n=2,270), a refined dataset of N=1,580 fully validated crash events were established as the final unit of analysis for the econometric modeling framework. The study employed a Standard Multinomial Logit (MNL) and a Mixed Logit (ML) framework, in order to account for unobserved heterogeneity. The empirical results establish that the Mixed Logit specification provides superior statistical goodness-of-fit, yielding a log-likelihood of -968.30, an Akaike Information Criterion (AIC) of 1,998.60, and a McFadden pseudo-R^2 of 0.279, compared to -1,042.60, 2,131.20, and 0.223 respectively for the standard MNL model. The study establishes that narrow lanes (lane width) (< 3.3 m), shoulder width (< 1.2 m), sharp horizontal curves (< 300 m), steep vertical grades (> 4%), and sight distance (< 120 m) increase the probability of fatal and severe truck accidents. The study further established the random nature of the variables including shoulder width, horizontal curve radius, and overloading, with recommendations for the need for improvements in the geometric design of the road, including widening of lanes, gentle horizontal curves, reduced steep grades for improved safety, the need for improved runaway truck ramps, and crash cushions, the use of GIS-based real-time monitoring of commercial vehicles, and the enforcement of axle-load regulations. The Sustainable Development Goals 3, 8, 9 and 11., is what this study aligns with, as it also provides practical information for geometric design optimization, corridor enforcement, and infrastructure investments.

Index Terms

Browse additional articles authored by one or more contributors to this paper.