Date of Completion


Embargo Period



Dr. Nalini RaviShankar, Dr. Karthik C. Konduri

Field of Study

Civil Engineering


Master of Science

Open Access

Campus Access


Pedestrian Conflict counts provides an important insight in predicting crashes and helps in identifying unsafe roads and intersections. This thesis work describes the estimation of pedestrian crash count and interaction severity prediction models for a sample of signalized intersections in Connecticut with either concurrent or exclusive pedestrian phasing. With concurrent phasing, pedestrians cross during the regular vehicle phase in the direction they are crossing the street, while with exclusive phasing, pedestrians cross in their own phase when all vehicles are stopped. Pedestrians crossing each intersection were observed and classified according to the severity of any interactions with motor vehicles. Intersections were selected to represent both types of signal phasing while controlling for other physical characteristics. In the nonlinear mixed models for interaction severity fixed effects of exposure measures and roadside characteristics, and random effects of intersection were entered nonlinearly to account for the repeated measures of pedestrian crossings at each intersection. For the models of interaction severity, partial proportional odds was found to fit the interaction data best, and annual average daily conflicts as pedestrian exposure formed the best model for predicting crash count on the basis of adequacy of model fit, significance of covariates, and goodness of fit tests. Intersections with exclusive pedestrian phasing have more minor or severe conflicts between pedestrians and motor vehicles but fewer crashes.

Major Advisor

Dr. John N. Ivan