Spatially-weighted Network Dual K function: Model Development and Application to Healthy and Unhealthy Food Environment

Keywords: network spatial analysis, point pattern analysis, Ripley’s K function, GIS
Abstract Type: Paper Abstract

Abstract

Capturing spatial co-location patterns—subsets of two or more groups of events that are geographically close—is one of the primary interests in spatial analysis because of its versatility. To reveal the spatial clustering of different types of point events on street networks using an exact statistical formula, the network dual K function method has been recently developed. However, the method has still some limitations. One of the major drawbacks is to assume a uniform distribution of events as a null hypothesis against which to test the degree of co-location. While this uniformity assumption is one of the most frequently encountered null hypotheses in spatial statistics, the actual distribution is quite likely to be uneven because of the biased density of consumers and urban establishments. To overcome this limitation, i.e., to take into account the inhomogeneous distribution, we propose an extended method, which is called the spatially-weighted network dual K function. The objectives of this paper are 1) to formulate an exact statistical method for analyzing co-location along streets in a central district constrained by a street network considering inhomogeneous distribution; 2) to implement this statistical method in computational procedures using free GIS software packages; 3) to apply the method to different types of healthy and unhealthy stores in a trendy district in Tokyo. The proposed method and its application are useful for detecting disadvantaged areas or groups of people with low accessibility to healthy food shops or being frequently exposed to unhealthy food shops.