Home Skip to main content

Project Details

Title:A Systems-Level Analysis of Left-Turning Vehicle-Pedestrian Crashes
Principal Investigators:Wesley Marshall and Bruce Janson
University:University of Colorado Denver
Grant #:69A3551747108 (FAST Act)
Project #:MPC-647
RiP #:01769397
RH Display ID:156614
Keywords:cities, crash data, left turns, pedestrian vehicle crashes, statistical analysis


Left-tuning vehicle-pedestrian crashes have long been dangerous for pedestrians in situations when drivers should be yielding the right of way. They outnumber right-turning vehicle-pedestrian crashes by a factor of 3 to 1 and are grossly overrepresented in terms of crash severity. If a driver does not properly yield, it is easy for traffic engineers to tally such crashes among the more than 90% of crashes that we attribute to human error. Yet, an accumulation of challenging conditions for a driver suggests that left-turning vehicle-pedestrian crashes are systematic problems and not random crashes caused by human error.

Accordingly, the proposed project seeks to take a system-level approach to studying this crash type via an empirical, macroscopic analysis of eight cities across multiple years. This includes: i) determining where this crash type is over- or under-represented while controlling for the level of pedestrian activity; and ii) statistically evaluating what combination of signal, design, and/or policy approaches associates with better or worse safety outcomes while also accounting for crash migration. Instead of focusing solely on signalization solutions, as is common in the existing literature, the intent is to shed light on how cities can employ a combination of approaches.

Project Word Files

NDSU Dept 2880P.O. Box 6050Fargo, ND 58108-6050