Week 10 Statistical Analysis with Network Data

In his book, John Scott (2012) commented on a struggle shared among many researchers – including many of us – who are attracted to SNA because of specific analytical needs but get turned away by SNA’s mathematical and statistical complexity.

One central goal of this course is to address this challenge, by focusing on essential SNA concepts – e.g., density, centrality, cliques, etc. – instead of the nitty-gritty mathematical details involved in SNA, so that we feel confident in designing an SNA study and venturing into directions made possible by SNA and network analysis in general.

So far in this course, you have been introduced to ways to generate descriptive measures of networks, or different types of network summaries, using the igraph package.

This week, we will stretch to statistical analysis with network data that may unleash the power of work you’ve been doing so far. In particular, we will work to:

  • Understand differences between the mathematical and statistical approaches to SNA
  • Become familiar with different kinds of statistical analysis with network data
  • Plan on applying statistical modeling approaches to your SNA project (if possible)

10.1 Introduction

The video at the bottom provides an introduction to this week.

Before this video, I want to mention that statistical models and modeling will get mentioned quite frequently in this week’s text. These terms are central to statistics in general. What a model does is to provide a statistical approximation of the ‘reality’ we are interested, so that by generating ‘better’ approximations we are essentially getting closer to truly capturing the ‘reality.’14 I personally like a point made by (Burnham and Anderson 2003) that all models are essentially wrong:

Fundamental to our paradigm is that none of the models considered as the basis for data analysis are the ‘true model’ that generates the biological data we observe… A model is a simplification or approximation of reality and hence will not reflect all of reality (p. 20)

This approach to considering statistical models might be useful as you approach this week’s readings.

10.2 Week 10 Activities

10.2.1 Read, Annotate, and Share

  1. Read Carolan (2014), ch. 8 and ch. 9, till the end of Regression.

  2. Annotate as we normally do using proper hashtags (e.g., question, idea, stats) and doing our ABCs (i.e., “Ask a question”, “Brag about your understanding”, and “Connect another peer’s ideas”). Build community knowledge: Because this week’s topic is of great complexity, we’re going to rely on our collective wisdom to develop deeper understanding. So please seek additional resources on your own, share back via Hypothesis or Slack, and help out a colleague.

  3. Finally, you can also find a study from education or business that deals with statistical analysis of network data, and share the study in the Slack general channel. Please point out specific techniques applied in that study (e.g., t-tests, ANOVA, regression, ERGMs).

10.2.2 Additional learning resources

If you are interested in going deeper in statistical analysis of networks, below are additional resources you can explore:

10.2.3 Course Project Check-in: Network analysis plan

I shared this course project rubric that you can follow while working on your course project.

This week, please build on your work from earlier weeks and craft a data analysis plan that can be added to your project artifact. If your data collection plan has changed, you can also provide an updated description of you network data. Share your plan as a Slack Post on the projects channel.

Please note each project is unique and let the community know if you are uncertain about any aspects of the rubric.

10.2.4 Class Meeting in Zoom

We will meet again in Zoom on Tuesday, Mar 30, 2021, 3–4:30pm Central Time. Please use the Zoom link in the calendar invite from me.

Have a great week!

References

Apkarian, Jacob, and Robert A Hanneman. 2016. Statistical Analysis of Social Networks. Jamaica, NY: City University of New York, York College.

Burnham, Kenneth P, and David R Anderson. 2003. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach. Springer Science & Business Media.

Kolaczyk, Eric D, and Gábor Csárdi. 2014. Statistical Analysis of Network Data with R: Use R! Springer New York. https://doi.org/10.1007/978-1-4939-0983-4.

Scott, John. 2012. Social Network Analysis. SAGE. https://doi.org/10.5040/9781849668187.

Snijders, Tom A B. 2011. “Statistical Models for Social Networks.” Annual Review of Sociology 37 (1): 131–53. https://doi.org/10.1146/annurev.soc.012809.102709.

  1. Note that the description of models introduced here may not fit the philosophical worldview you feel comfortable with or subscribe to. Refer back to Section @ref{threelevels} for an earlier discussion we had about aligning methodology and philosophical viewpoints.↩︎