Can System Dynamics Learn from Social Network Analysis?

Lukas Klaus Schoenenberger,Andrea Schenker-Wicki

doi:10.2139/ssrn.2550593

Abstract

This article deals with the analysis of large or complex system dynamics (SD) models, exploring the benefits of a multimethodological approach to model analysis. We compare model analysis results from SD and social network analysis (SNA) by deploying SNA techniques on a pertinent example from the SD literature — the world dynamics model. Although SNA is a clearly distinct method from SD in that it focuses on social actors and their interrelationships, we contend that SD can indeed learn from SNA, particularly in terms of model structure analysis. Our argumentation follows renowned system dynamicists who acknowledge the potential of SD to synthesize and advance theories in social science at both the conceptual and technical levels.

Highlights

This article deals with the analysis of large or complex system dynamics (SD) models, exploring the benefits of a mixing methods approach towards model analysis
social network analysis (SNA) is a clearly distinct method from SD in that it focuses on social actors and their interrelationships, we contend that SD can learn from SNA, in terms of model structure analysis
The remainder of this section is organized as follows: first, we give a short introduction into graph theoretical descriptions of directed networks and present the four centrality measures used in SNA; second, we transform Forrester’s world dynamics model into a digraph and calculate the four centrality measures for all nodes and third, we examine if the results achieved by SNA are valuable for SD modeling and analysis

Summary

Introduction

This article deals with the analysis of large or complex system dynamics (SD) models, exploring the benefits of a mixing methods approach towards model analysis. The most prominent example of this are the combinations of SD with cognitive mapping (Eden, 1994; Ackermann et al, 1997; Stotz and Größler, 2007), soft systems methodology (Lane and Oliva, 1994; 1998; Rodriguez-Ulloa and Paucar-Caceres, 2004; 2005), cybernetics (Schwaninger et al, 2004; Schwaninger and Pérez Ríos, 2008), and multicriteria analysis (Brans et al, 1998; Santos et al, 2002; Pruyt, 2007) All of these demonstrate the power and utility of a multimethodological approach. All of these demonstrate the power and utility of a multimethodological approach. Greene et al (2001, p.27) believe “that the fundamental uncertainty of scientific knowledge— especially about complex, multiply-determined, dynamic social phenomena—can be better addressed through the multiple perspectives of diverse methods than through the limited lens of just one.”

Objectives

Findings

Conclusion