TRB Pubsindex
Text Size:
x

« Prev Next »

Return to Results | New Search

Bookmark and Share

Title:

Analysis of Bicycle Crashes with Spatial Autocorrelation: A Comparison of Conditional Autoregressive Models

Accession Number:

01628227

Record Type:

Component

Abstract:

Bicycle crash frequency models are usually developed using macro-level crash data to evaluate the effects of various socioeconomic and demographic, land use, bicycle infrastructure, roadway, and traffic factors. One inherent property of macro-level or aggregate data is the presence of spatial correlation in the data, which is largely ignored in traditional negative binomial models. Conditional autoregressive (CAR) models within the hierarchical Bayesian framework are popular alternatives to account for spatial correlation in macro-level data. However, the scope of the CAR models is not broadly explored in transportation safety. This study estimates a number of different CAR models that take different structure for analyzing spatial dependence in the data. The analysis is performed using four years of bicycle crash data from 2011 to 2014 aggregated over census block groups in Florida. To gain more understanding of the relationship between bicycle crashes and bicycle trips, bicycle activity data collected from Strava smartphone application were used in the models. The results show that the presence of spatial correlation is significant in the bicycle crash frequency data, indicating bicycle crashes at neighboring census block groups is likely to be similar compared to those at distant census block groups. The variables that are credible within the Bayesian 95% credible interval and tend to increase bicycle crashes at census block groups include traffic volume, road density, urban principal arterials, urban collectors, occupied housing units, male population, younger population, households with no automobile, work trips by bicycle, and bicycle activity.

Supplemental Notes:

This paper was sponsored by TRB committee ANB20 Standing Committee on Safety Data, Analysis and Evaluation.

Monograph Accession #:

01618707

Report/Paper Numbers:

17-06880

Language:

English

Corporate Authors:

Transportation Research Board

500 Fifth Street, NW
Washington, DC 20001 United States

Authors:

Saha, Dibakar
Alluri, Priyanka
Wu, Wanyang
Gan, Albert

Pagination:

22p

Publication Date:

2017

Conference:

Transportation Research Board 96th Annual Meeting

Location: Washington DC, United States
Date: 2017-1-8 to 2017-1-12
Sponsors: Transportation Research Board

Media Type:

Digital/other

Features:

Figures; References (50) ; Tables

Geographic Terms:

Florida

Subject Areas:

Highways; Pedestrians and Bicyclists; Safety and Human Factors

Source Data:

Transportation Research Board Annual Meeting 2017 Paper #17-06880

Files:

TRIS, TRB, ATRI

Created Date:

Dec 8 2016 12:50PM