Introduction to Social Network Analysis

WELCOME!

This book is designed as an introduction to social network analysis. This is an exciting method, but it can also be very complex statistically and very challenging in terms of software coding. For many years, we have been teaching social network analysis across multiple levels of college coursework – from first semester college freshmen to advanced PhD students. We think that social network analysis can be easy to understand and apply. It does not require advanced statistical knowledge nor an extensive background in coding. Our goal is to break down the ideas and the procedures into an easily digestible way so that people who are learning about SNA for the first time can quickly understand and gain many useful skills.

Keeping it Simple

How do we accomplish this? By keeping things simple.

1. Rather than utilizing a dizzying array of different packages for conducting SNA, we instead rely almost entirely on igraph. That way, you can focus on working with one set of objects rather than having to move across multiple sets.

2. We make heavy use of network visualizations in order to illustrate the key concepts and procedures. We use, almost exclusively, one set of plotting tools: ggraph. This package is preferred over others because the coding logic is based on the widely used ggplot suite of visualizations, so working in ggraph will be easy for folks who are already familiar with ggplot and will offer highly transferable skills for newcomers.

3. We structure the book by first taking a deep dive into standard one mode, holistic networks. That way readers can keep this common form of network data in mind while working through the foundational concepts. Once this knowledge is established, we move on to extensions of the one mode framework, with demonstrations of working with and analyzing two-mode, ego-centric, text, and dynamic networks.

Web Applications and Datasets

Furthermore, we have created a series of web applications that can be used to practice the procedures presented in the book. No coding is required. Just point-and-click to recreate the visualizations and analyses. We have one app for each form of data and we have populated these apps with the example datasets presented in the book (as well as a few others that are not discussed).

Here are links to those applications…

• One mode network app

• Ego network app

• Two mode network app

• Text network app

• Dynamic network app

…and in the appendix you can find more information about the datasets.

• Network data descriptions

Advanced Study

While our main goal is to make SNA accessible to new learners, this book also has much value for more advanced students. In fact, much of the content included here has been used as part of the curriculum for our PhD classes on SNA. So while this book emphasizes the logic of SNA, it also comments on the statistics that underlie these procedures. We also point readers to advanced sources for further study.

Moreover, advanced students will benefit from recreating the analyses directly in R and RStudio. Throughout the book, we present and carefully annotate our code so that you can see and understand how we manipulate the data, generate the visualizations, and conduct the analyses. Our code, along with the data, can be used to reproduce everything that is shown here. You can check out the R and RStudio chapter in the appendix for details.

Acknowledgements

The adjoining web applications were expertly developed by Gaureth Sheth. Funding to pay for his time and effort was provided by an Alt-Textbook Grant from the North Carolina State University Libraries. Years earlier, the lead author (McDonald) collaborated with Dr. Jim Yocom on the development of a prototype Shiny app for SNA instruction. That app, which is still in use today, served as the inspiration for writing this textbook.

This project also drew inspiration from our longstanding NC State working group on social network analysis – WolfWebs. Through generous seed funding from NC State’s Data Science and AI Academy (DSA), we have been able to expand the group across disciplines and to support a multi-year speaker series on SNA. The DSA also supported Leppard on a postdoctoral fellowship while this textbook was under construction. We are grateful to the pack of WolfWebs affiliates for motivating us to write this textbook. Two individuals deserve specific mention. Dr. Robin Dodsworth is a founding member of WolfWebs and active in our community. Adam Goldfarb also played a central role in creating and sustaining the group. Much thanks to you both.