Network paradigms are on the rise. Over the past two decades, they have diffused from sociology, anthropology, and social psychology into count-less other disciplines. More and more phenomena are being conceptual-ized as networks because it appears that they are better suited to representing and explaining some complex dependencies that are not captured in population samples and feature vectors.
Accordingly, an increasing number of books about networks and net-work analysis is also being published. In our view, however, these are often centered on substantive problems addressed using network perspec-tives, on theoretical reflections about the role and function of networks, or on models and methodological contributions for the analysis of net-works and complex systems in general. What appears to be missing is a textbook that builds on and updates the widely known and instantly classical texts by Wasserman and Faust (1994) and Scott (2000) but is more directly tailored to application in empirical studies.
The idea to put together an interdisciplinary group of researchers to address this apparent gap arose in a discussion between Marina Hennig and Ines Mergel during the Sunbelt XXVI Social Networks Conference (25-30 April 2006, Vancouver, BC). Steve Borgatti, Ulrik Brandes, Marina Hennig, Lothar Krempel, Ines Mergel, and Michael Schnegg then outlined a structure that would follow the empirical research process and include learning goals, examples, and exercises. This initial design was refined and implemented by the present authors. An important aspect of our approach is that we do not focus on networks in a specific domain or methods of a particular kind; instead, our intention is to provide a compact guide to the utilization of network approaches in the social sciences, in the broadest possible sense.
Many of these aspects have been shaped by the valuable feedback we received from participants in courses we have taught in various formats, disciplines, and locations. Additional input was provided by numerous colleagues and students within and outside of our working groups. We have also learned from each other, and sincerely hope that readers will feel that they benefit from this book as much as we did from the experience writing it.
Our heartfelt thanks go to everyone who has contributed to the com-pletion of this project. The Deutsche Forschungsgemeinschaft (DFG) supported us financially with a Cooperation Network grant to Marina Hennig and a Reinhart Koselleck project grant to Ulrik Brandes. The Social Science Research Center Berlin provided administrative management support, and we arc particularly grateful to Elisabeth Hamacher, Susanne Grasow, and Claudia Buchmann. We thank Susan Cox for copy-editing and Christine Agorastos for assistance in preparing the final manuscript. Jutta Allmendinger and Kathleen M. Carley provided much-appreciated guidance and support, and our own personal networks bore with us.

Preface 9

How to Use this Book 11

1 Introduction 13

1.1 The Construction of Social Networks 13

1.2 Social Network Studies 15

1.2. 1The Community Question 15

      1.2.2 Viral Marketing 20

      1.2.3 Corporate Networks 21

1.3 Exercises 25

2 Research Design 27

2.1 Social Networks 29

2.2 Networks as Variables 31

      2.2.1 Explanatory Variables 35

      2.2.2 Dependent Variables 37

2.3 Typology of Networks 47

      2.3.1 Complete Networks 49

      2.3.2 Ego-centered Networks 52

2.4 Longitudinal Network Studies 55

2.5 Summary 57

2.6 Exercises 59

3 Data 61

3.1 Kinds of Data 62

      3.1.1 Units and Levels 62

      3.1.2 Organization 67

      3.1.3 Which Data for Which Type of Network? 72

3.2 Data Collection 75

      3.2.1 Sources 75

      3.2.2 Boundary Specification 83

      3.2.3 Alter Recall 85

3.3 Quality Issues 93

3.4 Ethical Considerations 97

3.5 Summary 99

3.6 Exercises 100

4 Analysis 103

4.1 Mathematical Representation 104

      4.1.1 Graphs 106

      4.1.2 Ego-Centered Networks 109

      4.1.3 Two-Mode Networks 110

4.2 Indexing and Grouping 111

      4.2.1 Dyads as the Unit of Analysis 112

      4.2.2 Network Characteristics 118

      4.2.3 Centrality 123

      4.2.4 Cohesion 130

      4.2.5 Roles 134

      4.2.6 Blockmodeling 137

4.3 Modeling 140

      4.3.1 Idealized Models 141

      4.3.2 Exponential-Family Random Graph Models 143

4.4 Summary 145

4.5 Exercises 146

5 Visualization 149

5.1 Graphical Representation 151

      5.1.1 Sociogram 151

      5.1.2 Sociomatrix 155

      5.1.3 Two-Mode Network Representations 159

5.2 Multivariate Information Visualization 162

      5.2.1 Substance-Based Layout 165

      5.2.2 Other Graphical Variables 169

5.3 Information Layering 169

      5.3.1 Filtering 170

      5.3.2 Level of Detail 174

      5.3.3 Micro/Macro Reading 175

5.4 Summary 177

5.5 Exercises 179

6 Summary 183

List of Figures 189

List of Boxes 191

About the Authors 193

Bibliography 195

Index 211

Marina Hennig is Professor for Social Network Research and Sociology of the Family at Johannes Gutenberg-University Mainz. She studied so-ciology at Humboldt Universität zu Berlin. She was awarded her doctorate in 1999 and venia legendi in 2006, both by Humboldt Universität. Her post-doctoral thesis was about individuals and their social relations and her research focuses on ego-centered network analysis, family ties in contemporary urban Germany, the generation of social capital in online social networks, and the further substantiation of the network perspective. Her most recent research project (2009-2011) focused on an empirical reconstruction of the Pierre Bourdieu's habitus-field theory of using network analysis. PA\Ulrik Brandes has been Professor for Algorithmics at the University of Konstanz since 2003. He studied computer science at RWTH Aachen and received his doctorate degree and venia legendi in computer science from the University of Konstanz in 1999 and 2002, respectively. His research interests include network analysis, graph algorithms, experimental algorithmics, graph drawing, and information visualization. He has been a member of the Board of Directors of the International Network for Social Networks Analysis (INSNA) since 2008. Jürgen Pfeffer earned a PhD in Business Informatics from Vienna University of Technology, Austria. He worked in industry and non-university research institutes for ten years and is currently a postdoctoral associate at the School of Computer Science at Carnegie Mellon University. Jürgen Pfeffer's research combines traditional network analysis and dynamic network analysis theories and methods with up-to-date science from the areas of visual analytics, geographic information systems, system dynam-ics, and data mining. His research focus lies in the computational analysis of organizations and societies with a special emphasis on dynamics and change in large-scale systems. Ines Mergel is Assistant Professor of Public Administration and Interna-tional Affairs at the Maxwell School of Citizenship and Public Affairs, Syracuse University, NY. She studied business administration at the Uni-versity of Kassel, Germany, and the Reijksuniversiteit Leiden, The Neth-erlands. She was awarded a doctorate in business administration by the University of St. Gallen, Switzerland, in 2005 and spent six years as a research fellow at the Program on Networked Governance at Harvard Kennedy School of Government, Cambridge, MA. Her research focuses on the diffusion and adoption of social media applications among public managers in the U.S. federal government. She is the author of Social Me. dia in the Public Sector: A Guide to Participation, Collaboration and Trans-parency in The Networked World (2012) and co-authored the companion field guide (2012, with Bill Greeves). Her research has been published in leading public administration journals.

中科院文献情报中心