Determining causality and dependency in loosely coupled,n-dimensional social networks

Abstract

Social networks permeate our lives. The desire to understand them pervades much of social science today. This chapter offers an empirically sound method for analyzing causal and dependency relationships among the people, places, things, and concepts that flow within and between social networks. A particular emphasis of this approach is modeling and analyzing the connections between social networks and the physical networks that enable social networks. Most social networks lack a fixed organizing principle or any discernable, formal structure. This results in a loose coupling of elements within or between networks. This also makes identification of boundary layers difficult. Further, the dimensionality of loosely coupled networks can grow enormously. The approach described here, called Williamsburg, addresses three issues in social network analysis: loose coupling of networks, dimensionality, and the need to test empirically the analytic findings from our approach to social network analysis.