Analyzing injury severity of three-wheeler motorized rickshaws: A correlated random parameters approach with heterogeneity in means

• Published in: Accident analysis and prevention Vol. 204; p. 107651
• Main Authors: Pervez, Amjad, Jamal, Arshad, Haider Khan, Salman
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.09.2024
• Subjects: Correlated random parameters model; Developing country; Injury severity risk factors; Three-wheeler motorized rickshaws; Unobserved heterogeneity; Injury severity risk factors; Three-wheeler motorized rickshaws; Correlated random parameters model; Unobserved heterogeneity; Developing country
• ISSN: 0001-4575; 1879-2057; 1879-2057
• DOI: 10.1016/j.aap.2024.107651

•Paper seeks to explore significant injury severity risk factors in 3-WMR collisions.•Correlated random parameter modeling framework is adopted to address the unobserved heterogeneity.•Importance of considering unobserved heterogeneity is demonstrated.•Study provides useful insights and policy implications to improve the safety of 3-WMR riders. Traffic crashes involving three-wheeler motorized rickshaw (3-WMR) are alarming public health and socioeconomic concerns in developing countries. While most of the earlier studies have dealt with safety analysis of four- and two-wheelers, there is a noticeable gap in understanding the safety dynamics, especially the risk factors affecting the crashes involving 3-WMR. The present study aims to address this gap by exploring potential risk factors influencing 3-WMR crashes, utilizing a correlated random parameters multinomial logit model with heterogeneity in means (CRPMNLMHM). This modeling framework advances the classic random parameters model by capturing associations among random parameters, providing a more comprehensive understanding of crash risks associated with 3-WMR. The empirical analysis draws on three years of traffic crash records (2017–2019) maintained by RESCUE 1122 in Rawalpindi city, Pakistan. A comparative assessment between the modeling frameworks demonstrated that CRPMNLMHM outperformed its counterparts. Model assessment for heterogeneity in the means identifies two significant variables, i.e., young age and nighttime, which yield statistically significant random parameters. In addition, the model's results suggest that fatal and severe injury outcomes in 3-WMR crashes are affected by several attributes related to temporal characteristics (weekend, nighttime, and off-peak indicators), driver profiles (young, older aged, and speeding), posted speed limits (>70 kmph), weather conditions (raining), and crash characteristics (collision with pedestrians, trucks, or 3-WMR overturning). The present study’s findings offer invaluable insights, emphasizing the significance of considering for unobserved heterogeneity in variables contributing to the injury severity of 3-WMR crashes. Moreover, in light of the findings, a set of policy implications are suggested, which will guide safety practitioners to develop more effective countermeasures to address safety issues associated with 3-WMRs.