bicycle-safety-roundabouts

MPC-603 Investigating bicyclist safety perceptions and behaviors at roundabouts

This repository contains data and scripts associated with the "Investigating bicyclist safety perceptions and behaviors at roundabouts" research project MPC-603, funded by the Mountain-Plains Consortium, and carried out by Patrick Singleton and the Singleton Transportation Lab at Utah State University.

• Objectives
  • Characterize bicyclists’ safety perceptions of roundabouts and roundabout elements.
  • Identify bicyclists’ preferences for various roundabout elements.
• Methods
  • Survey data collection, using an online questionnaire about bicyclists' personal characteristics, user behaviors, perceptions of comfort, and preferences for various design and operational elements of roundabouts.
  • Statistical modeling of associations between safety perceptions and user characteristics, roundabout designs, and other factors.

Publications

• Poudel, N., & Singleton P. A. (2021). Bicycle safety at roundabouts: A systematic literature review. Transport Reviews, 1877207. https://doi.org/10.1080/01441647.2021.1877207
• Poudel, N., & Singleton, P. A. (2022). Preferences for roundabout attributes among US bicyclists: A discrete choice experiment. Transportation Research Part A: Policy and Practice, 155, 316-329. https://doi.org/10.1016/j.tra.2021.11.023

Description of files and folders

• Many of these scripts were written in R. To use, download R and then download RStudio.
• Many of these scripts were written in Python to use with Biogeme. To use, follow installation instructions.
• bicycle-safety-roundabouts.Rproj: This RStudio project file allows R users to use relative file paths from this root folder, if R scripts are opened after opening this R project file.

Data

These folders contain data collected, assembled, and generated as part of this research project.

• Census: Files associated with weighting the data.
• Experimental designs: Files associated with generating and selecting the experimental designs.
• Final images: JPG and GIF files depicting the photorealistic simulated roundabouts.
• Survey: Anonymized survey data collected, along with questionnaire.

Analysis

These folders contain processed data, scripts, and outputs associated with analyses of the research project data.

• Data cleaning: Files associated with the processing and cleaning of the survey data.
• Stated preference: Files associated with the discrete choice analysis of stated preference survey data.
  • Data cleaning and weighting: Files associated with the preparation of the survey/experimental data for the analyses.
  • Heterogeneity model: Files associated with the panel mixed multinomial logit models, including systematic preference heterogeneity.
  • MMNL model: Files associated with the panel mixed multinomial logit models.
  • MNL model: Files associated with the multinomial logit models.
• Comfort analysis: Files associated with the analysis of perceived comfort data.
  • Data cleaning: Files associated with the preparation of survey data for the analysis.
  • Comfort analysis: Files associated with the multivariate ordered probit models of comfort.