Modeling pedestrian injury severity in pedestrian-vehicle crashes considering different land use patterns: Mixed logit approach

Abstract

Objective

The objective of this study is to identify and compare the contributing factors to pedestrian injury severity in pedestrian-vehicle crashes considering different land use patterns.

Methods

The pedestrian-vehicle crash data from 2007 to 2018 were collected from the North Carolina Department of Transportation (NCDOT). A total number of 15,807 observations with 72 categorical variables were included in the final dataset. Two mixed logit models were developed to analyze the crash dataset with segmentations of two dominant land use areas (i.e., residential and commercial). Fixed and random parameters were found in both models. Estimation results and marginal effects of significant explanatory variables were investigated.

Results

In general, the residential model has 24 fixed parameters and 3 random parameters. The commercial model has 31 fixed parameters and 3 random parameters. According to the estimated results, elder or drunk factors are found to have more impacts on severe injuries in residential areas. Large and mid-size vehicles increase the probability of severe injuries in commercial areas. The marginal effect values for severe injury at non-intersections have opposite signs in the two models. Besides, speed limits between 40 and 45 mph and factors related to poor visibility are more likely to result in severe pedestrian injuries. Coarse asphalt pavement can reduce the probability of severe pedestrian injuries.

Conclusions

This study investigated the pedestrian injury severity in pedestrian-vehicle crashes considering two types of land use using a mixed logit approach. Based on the discussions of factors contributing to the pedestrian injury severity, policies and countermeasures to improve traffic safety are suggested. Above all, a mixed-use land development policy is recommended. Other suggestions are summarized below: (1) giving more considerations to older pedestrians when planning and designing residential areas; (2) strengthening laws and education against drunk driving and even drunk walking on/across the roadways; (3) increasing the frequency of the patrols and alcohol tests; (4) improving lighting conditions and road alignments; (5) establishing a limited-truck-passing-period policy especially in commercial areas; and (6) improving the pavement conditions wherever needed.

Keywords:

• Pedestrian-vehicle crashes
• pedestrian injury severity
• mixed logit model
• land use

Additional information

Funding

The authors want to express their deepest gratitude to the financial support by the United States Department of Transportation, University Transportation Center through the Center for Advanced Multimodal Mobility Solutions and Education (CAMMSE) at The University of North Carolina at Charlotte (Grant Number: 69A3551747133).