Estimating Pedestrian Exposure Prediction Model in Rural Areas

Abstract

"Pedestrian exposure" is defined as the exposure risk of pedestrians to collisions with motor vehicles. It is one of the important factors influencing pedestrian crashes. Because pedestrian exposure or even pedestrian volume counts are not readily available, population density is usually used as a substitute in pedestrian crash prediction models. Unfortunately, population density is not a good replacement for pedestrian exposure because it does not account for the amount of walking people do. This study investigates the relationship between the weekly pedestrian exposure in rural areas of Connecticut and factors such as population density, presence of sidewalks, number of lanes, area type, traffic control type, and median household income. General linear modeling and Tukey and Duncan multiple comparison of means methods are used to identify the significant factors. Only the number of lanes, area type, and sidewalk system significantly explain the variation in the resulting pedestrian exposure prediction model. This study suggests extra improvement in pedestrian facilities for the areas with high pedestrian exposure. Ongoing research will take advantage of the model to estimate pedestrian crash models in rural areas of New England.