Safe-D: Safety through Disruption

Sources and Mitigation of Bias in Big Data for Transportation Safety

Abstract

High quality data for transportation safety planning has been expensive and slow to obtain. Recently, new big data sources allow more detailed analysis of vehicle, transit, bicycle, and pedestrian trips than ever before. However, big data generally represents transactions rather than trips–inherently including a range of biases related to representation. Big data sources offer both prospect and problems for transportation planning, regarding how well they reflect the broad population of transportation system users, or individual markets subject to digital divide and other biases of representation. Research has identified far-reaching bias issues in big data sources, but this study will focus on those with an impact to planning for transportation safety. Using a synthetic literature review, and interviews with expert practitioners, Results suggest implications for transportation safety research and practice to identify and mitigate bias in big data.

Highlights

  • Researchers conducted a qualitative review of literature (n=75) and semi-structured interviews with big data experts (n=10).
  • Interviews coded by two independent researchers were moderately reliable (multi-valued nominal alpha = 0.544).
  • Interviewees most often referred to big data for transportation safety concerning vehicles, followed by bicycling and transit.
  • Study identified four categories of bias related to big data for transportation safety, and aligned strategies to mitigate bias in research and practice.

Final Report

02-026 Final Research Report (PDF)

EWD & T2 Products

Webinar (ppt): Griffin, G. (2018, June 26). “What do the Experts Do? Insights from Interviews & Literature to Deal with Bias in Big Data.” In Conversations about Counting: Big Data – Implications for Bicycle and Pedestrian Traffic Analysis, Webinar Presentation to the Transportation Research Board Bicycle and Pedestrian Data Subcommittee.

Student Impact Statement (pdf): Two students were funded under this project (PhD student Greg Griffin from TTI and Master’s student Meg Mulhall from UT). This file contains a statement of the impact this project made on these students’ education and workforce development.

Presentations/Publications

Griffin, G. P., Mulhall, M., Simek, C., and Riggs, W. W. (2019) Mitigating Bias in Big Data for Transportation. Proceedings of the Transportation Research Board 98th Annual Meeting, No. 19-03196. Washington, D.C., Transportation Research Board. (Presented)

Thesis/Dissertations

This project contributed to the following master’s thesis or Ph.D. dissertations:

  • Mulhall, Megan (not yet published). A guide to big data types and applications for the transportation planner. Master’s Thesis. Master of Science in Community & Regional Planning, Texas A&M University.

Final Dataset

The final dataset for this project is located in the Safe-D Collection on the VTTI Dataverse; DOI: 10.15787/VTT1/KRTX66

Research Investigators (PI*)

Greg Griffin (TTI/TAMU)*

Project Information

Start Date: 2017-08-01
End Date: 2018-05-31
Status: Active
Grant Number: 69A3551747115
Total Funding: $21,761
Source Organization: Safe-D National UTC
Project Number: 02-026

Safe-D Theme Areas

Big Data Analytics

Safe-D Application Areas

Planning for Safety

More Information

RiP URL
UTC Project Information Form

Sponsor Organization

Office of the Assistant Secretary for Research and Technology
University Transportation Centers Program
Department of Transportation
Washington, DC 20590 United States

Performing Organization

Texas A&M University
Texas A&M Transportation Institute
3135 TAMU
College Station, Texas 77843-3135
USA