Parameterizing the Yielding Compliance of Motorists at Signalized Midblock Crosswalks Using Mixed Effects Logistic Regression
Document Type
Conference Proceeding
Publication Date
11-1-2017
Publication Title
Transportation Research Board 97th Annual Meeting
Abstract
This study evaluates the motorists yielding compliance at signalized midblock Danish offset crosswalk by incorporating the non-motorist traits, crosswalk characteristics, and temporal factors. The study utilized data from five crosswalks located in Las Vegas, Nevada, equipped with either Rectangular Rapid Flashing Beacon (RRFB), Overhead Flashing Beacon (OFB), or Circular Rapid Flashing Beacon (CRFB). The descriptive analysis shows that , in overall, the yielding compliance was 64.6%. However, when the flashes were active this yielding compliance increased to 77.1% with a varying rate (60% - 84%) across the sites. The Mixed Effects Logistic Regression (MELR) that considers the heterogeneity characteristics of the crosswalks’ treatment types and land use was then applied. The MELR results revealed that, motorists were 6 times more likely to yield when non-motorists arrived at the crosswalk while the flashes were already active and 3 times more likely to yield when a pedestrian pressed the button while flashes were inactive. Moreover, motorists were 75% and 33% less likely to yield to pedestrians crossing between the yield line and advanced pedestrian crosswalk sign and, yield line and marked zebra stripes respectively, and 15% less likely to yield to black non-motorists. Additionally, the higher the number of pedestrian crossing at a time the higher the odds of yielding. Lastly, the intersection green light downstream the crosswalk, negatively affected the yielding tendencies. Acting on the findings from this study would substantially improve non-motorists’ safety at the crosswalks.
Language
eng
Repository Citation
Kutela, B.,
Teng, H.
(2017).
Parameterizing the Yielding Compliance of Motorists at Signalized Midblock Crosswalks Using Mixed Effects Logistic Regression.
Transportation Research Board 97th Annual Meeting
COinS