A comparative study of safe and unsafe signalized intersections from the view point of pedestrian behavior and perception

Abstract

Signalized intersections with marked crosswalks enhance pedestrian safety by providing the designated right of way to pedestrians. However, a significant number of pedestrian fatalities occur at signalized intersections, which may primarily be due to pedestrians’ violation behaviors. Since pedestrians’ fatalities are not uniform across signalized intersections in a city, it may be expected that violations would also vary across the sites. It is thus worthwhile to investigate if the pedestrian signal violation is a good surrogate for fatal pedestrian crashes at signalized intersections, and if so, what behavioral, spatial, and built environment related factors influence such violations. To this end, present study analyzes pedestrian behavior and perceptions across twenty-four signalized intersections in Kolkata city, India, out of which twelve intersections did not record any fatal pedestrian crashes between 2011 and 2016 and the remaining twelve experienced at least three or more fatal pedestrian crashes over the same period.Using data from the video-graphic survey at these twenty-four signalized intersections violation behaviors are extracted along with personal attributes at the pedestrian level. Further, pedestrian perception surveys are carried out at each of the twenty-four sites, to obtain a user’s perception of safety and satisfaction. Results indicate that pedestrians’ signal violations behavior and dissatisfaction are statistically significantly higher at locations with recorded fatal pedestrian crashes. Results from different models and analysis clearly pointed out several planning and design deficiencies such as longer waiting time before crossing, higher pedestrian-vehicular interaction, pedestrian’s state of crossing, and a number of personal level attributes such as pedestrian’s intended mode of transportation and their state of journey, pedestrian’s home location, pedestrian’s socio-demographic characteristics as important predictors of pedestrians’ violation behavior. The methodology and findings are useful not only for proactive safety improvement at signalized intersections but also to proactively identify potential unsafe sites.