Review: John Scott, Social Network Analysis (2000)

I recently picked up a copy of John Scott’s Social Network Analysis: A Handbook when I was researching methodologies I could use in my study of Wikipedia and social networks. Scott, a sociologist, provides both a history of the development of social network analysis and an introduction to the basic terminology and concepts related to the discipline. While this second edition doesn’t seem to have been updated to account for the developments in the field since the first edition of 1991, I found it to be a more than adequate overview of the concerns and methodology of social network analysis.

(I have to provide a caveat here: I am not a sociologist and I’m completely new to social network analysis, so any criticisms—or praise—of this book on my part may be completely off-base. For example, I have no idea how social network analysis has developed over the last 15 years, so I’m not really competent to judge if Scott has left out any important additions or modifications to the theory and methodological practices used in social network analysis. My comment above is based solely on my impression in reading the text that the majority of Scott’s references in the text are pre-1990. Anyway, you should take most of my claims in this post with a grain of salt.)

Overview
Scott’s book is organized into three main sections: the first and second chapters outline the history of social network theory, while chapters 3–8 introduce basic terms and the methods of network analysis. Finally, an appendix lists software tools for conducting social network analysis with short reviews of each software package.

In chapter 2, Scott outlines the history of social network theory. According to him, it began with the focus on societal structures in the work of anthropologist Alfred Radcliffe-Brown during the 1920s and 1930s. Later, researchers, particularly a group centered at Harvard, combined elements of gestalt theory and the mathematical tools of graph theory to analyze these structures. One of the chief developments of this research was the formalization of the theoretical principles of network analysis, which helped to determine the basic methodology of social network analysis, and the development of the sociogram, a digram of the nodes and links in a network.

The remainder of the text serves as an introduction to social network analysis and the terminology and best practices used by network researchers. According to Scott, analyzing social networks is primarily a process of collecting and storing data. However, that data isn’t attributive data about a subject. Rather, it is relational data; that is, data about the connections between subjects. For this reason, when researchers collect social networking data, they should focus on these relations. The primary method of doing this is with incidence or adjacency matrices, where the former record binary information about the existence of a connection between two classes subjects and the latter record information about the number of connections between the members of a particular class. Further, these matrices can be directional, indicating that the connections between subjects do not necessarily flow both ways.

The analysis of social networks can be conducted using positional or reputational approaches. When using the positional approach, researchers are interested in investigating the social position that a particular subject occupies, whereas with the reputational approach, which is used primarily when there are no stable positions to investigate, researchers have their subjects suggest other subjects, and the connections between those subjects are mapped.

Final thoughts
Throughout the text, Scott emphasizes the difficulty that many researchers have with social network analysis because of its foundation in matrix algebra. However, there isn’t much math in the book, and what math is there is explained in an understandable manner. He acknowledges that most researchers will rely on software tools to do the number crunching for them, so he spends more time explaining the research rationale and grounds for using particular approaches to the study of social networks, emphasizing that individual projects will require different approaches and that researchers should understand what their tools are measuring, even if they aren’t sure how they are measured. As a rhetorician, I found Scott’s repeated insistence that researchers have a clear idea of their research goals before choosing the tools they will use so as to make sure that those tools are the best for their particular design to be refreshing, and it was easy for me to see how this type of analysis would be a good fit for ecological studies of social phenomena.

Overall, I found the book to be an accessible and readable introduction to the techniques of social network analysis.