Scale effects on spatially embedded contact networks

TL;DR

This study investigates how scale changes affect the network and spatial structures of spatially embedded contact networks, revealing different sensitivities and interactions that inform targeted health interventions.

Contribution

It provides a novel analysis of scale effects on both network and spatial structures, highlighting their interactions and scale sensitivities in contact networks.

Findings

Network properties are sensitive to scale at fine levels.

Spatial structure remains scale-independent.

Network behaviors are influenced by spatial structure.

Abstract

Spatial phenomena are subject to scale effects, but there are rarely studies addressing such effects on spatially embedded contact networks. There are two types of structure in these networks, network structure and spatial structure. The network structure has been actively studied. The spatial structure of these networks has received attention only in recent years. Certainly little is known whether the two structures respond to each other. This study examines the scale effects, in terms of spatial extent, on the network structure and the spatial structure of spatially embedded contact networks. Two issues are explored, how the two types of structures change in response to scale changes, and the range of the scale effects. Two sets of areal units, regular grids with 24 different levels of spatial extent and census units of three levels of spatial extent, are used to divide one observed and two reference random networks into multiple scales. Six metrics are used to represent the two structures. Results show different scale effects. In terms of the network structure, the properties of the observed network are sensitive to scale changes at fine scales. In comparison, the clustered spatial structure of the network is scale independent. The behaviors of the network structure are affected by the spatial structure. This information helps identify vulnerable households and communities to health risks and helps deploy intervention strategies to spatially targeted areas.