The Effect of the Built Environment on Pedestrian Volume in Microscopic Space - Focusing on the Comparison Between OLS (Ordinary Least Square) and Poisson Regression

Abstract

This study is aimed at establishing a correlation between microscopic factors and pedestrian volume in the urban environment, focusing on microscopic factors that stimulate pedestrian volume, such as density, diversity, network structure, accessibility, the form of lots and buildings, and the form of building facades. In particular, factors already known to boost pedestrian volume include density, diversity and accessibility, which are three variables highly related to the concept of 3Ds (Density, Diversity, Design) proposed by Cervero and Kockelman (1997) and the additional 2Ds (Distance to Transit, Destination Accessibility) suggested by Ewing et al. (2008). The analysis in this study is based on the 2010 survey of the floating population in Seoul, particularly on the data from Seocho-gu in the Gangnam area. Data was established by analyzing microscopic factors within the 500m radius around each of the 616 spots from which the pedestrian volume in Seocho-gu was measured. This study compared and analyzed two methods: OLS, which is featured in multiple studies of pedestrian volume, and Poisson Regression, which is the most common statistical analysis method of abnormal count data such as pedestrian volume. The analysis results showed that density, diversity and accessibility, three factors that were already known to be effective in increasing pedestrian volume, also proved to have the same effect in the Gangnam area. Moreover, the form of the ground level and facade of buildings were found to have a significant effect on pedestrian volume. These findings are expected to serve as basic data for the development of sustainable and resilient cities through higher pedestrian volume.