An Observational Evaluation of Pedestrian Behavior at Signalized Intersections

Pedestrians’ appropriate use of pedestrian infrastructure at signalized intersections—in particular, their compliance with the pedestrian signal and pushbutton use—can have important safety and efficiency implications. Therefore, it is important to understand why pedestrians do not always use infrastructure as engineers intend. However, there is a lack of research on pedestrian pushbutton use and pedestrian signal compliance, particularly at US locations. This research addresses this knowledge gap with an observational study of eight corners across four signalized intersections in Corvallis, Oregon, USA, where video data was collected and manually reduced to yield 2,633 unique pedestrian observations. At locations where pushbutton actuation was required to yield the WALK signal, over 80% of pedestrians traveling alone and arriving first at the corner pressed a pushbutton. For pedestrians who arrived on the DON’T WALK interval, results from a logistic regression model suggest that presence of conflicting vehicles on arrival, pedestrian waiting time, whether the pedestrian was first to arrive at the intersection corner on a cycle, and pedestrian pushbutton presence/use were significant factors in predicting the likelihood of the pedestrian entering on the DON’T WALK interval. These findings will help transportation practitioners better understand why—and the extent to which—pedestrians engage in “non-compliant” behaviors at signalized intersections and how engineering design choices can encourage compliance.