Citation

Saha B, Fatmi MR, Rahman MM. Transp. Dev. Econ. 2021; 7(2): 23.

Copyright

(Copyright © 2021, Holtzbrinck Springer Nature Publishing Group)

DOI

10.1007/s40890-021-00132-5

PMID

unavailable

Abstract

This study investigates pedestrian injury severity for Dhaka, the capital of Bangladesh--a developing country. A latent segmentation-based ordered logit (LSOL) model is developed utilizing the police-reported collision records of Dhaka for 2011-2015. Injury severity is modelled in the following three-point ordered scale of: (1) minor injury; (2) major injury; and (3) fatal. The LSOL model is developed to address the ordinal nature of injury severity levels, as well as capture heterogeneity by endogenously distributing pedestrians into discrete latent segments. This study tests the influence of built environment characteristics, such as road network configuration, land use, and transportation infrastructure attributes. The LSOL model is estimated for two segments. Model results suggest that segment one can be identified as a high-risk segment, whereas segment two is a low-risk segment. Parameter estimation results reveal that higher land use mix index, three-way and four-way intersections, and mid-block road segments aggravate the pedestrian's injury. The model confirms that significant heterogeneity exists across the segments. For instance, collisions occurring at the traffic police-controlled intersections in Dhaka are more likely to yield severe pedestrian injury in the high-risk segment. In contrast, the same variable shows a lower likelihood for severe injury in the low-risk segment. The elasticity effect analysis suggests that length of the sidewalk, distance to the railway line, and mid-block road segments show a substantial positive impact on fatality. The findings of this study will assist engineers and planners to develop plans and policies for improving pedestrians' safety in developing countries.

Language: en