Communication and Social Networks (Fall 2023)/Final project

From CommunityData

You can choose a final project from one of the following options:

  1. Network Intervention
  2. Network Analysis
  3. Company Pitch

More information about each option is below. For each project, you will produce a research report of some kind. You should write this in a Quarto file, and then export it to a Word document.

Please turn in two files:

  • Your .qmd file that includes your code
  • The Word document or PDF version of the report. Figure out to set echo=FALSE, so that it doesn't include all of your code, and just looks like a nice report.


Network Intervention[edit]

In teams of 3 or 4, you will design, carry out, and measure a network intervention, using principles from this class, like diffusion, contagion, social capital, etc. You will identify a message/product/idea/etc and try to spread it, either within or beyond Purdue.

I would encourage you to identify a goal that is meaningful to your group; projects should be creative, courageous, and challenging. We will evaluate your work not based on whether your are ultimately successful, but how innovate your approach is, and how grounded it is in the network theories we have learned about.

Examples[edit]

Brooke Foucault-Welles has run a similar project in her class. Examples of well-designed projects in her class include:

  • A comparison of the diffusion on Twitter of a plain-text and image-and-text message about feminism.
  • An alteration intervention that aimed to boost morale in a student organization by creating cross-cohort (year in school) friendship ties.
  • A marketing campaign that leveraged the friendship paradox to spread the word about a new student group.
  • An intervention that targeted student leaders to lobby University Health and Counseling Services to bring therapy animals to campus during finals week.

Presentation[edit]

Because this is a group project, I'll expect a lot more than the individual projects. Part of this is in being ambitious in the project that you select, but also in how you present your results.

In particular, those choosing this option will do a short (5 minute) presentation to the class at the end of the semester, to tell us about your project and your findings. It would be fun for those in the class to learn about what you've been doing.

Executive Report[edit]

The major part of the assignment, though, is a 5-6 page executive report. This report should have the following sections:

  • Introduction
    • In the introduction you will describe the motivation for your intervention. You will explain what you did, and why you chose that approach.
  • Network Theory
    • In your introduction, you will likely mention some network theories that helped to motivate your intervention. In this section, you will provide more detail about the theories that guide your work. You should include at least three references to support your approach and analyses.
  • Approach
    • In the introduction you will give a high-level explanation of your intervention. Here you will provide a more detailed description, together with a more extended argument about why this is a good approach to meet your goals.
  • Analysis
    • You will analyze the effect of your intervention. The data to do this can take different forms - e.g., you may survey people to learn about awareness or beliefs, or you might gather data from the web or gather network data. At some point in your report you should include 1) a network analysis and 2) at least 1-2 network visualizations. It's possible that these may come earlier. For example, you might use existing data or simulated data to explain a network theory or to motivate your approach. If you are able to gather network data, then the network analysis and visualizations would probably be part of this Analysis section.
  • Conclusion
    • Finally, write a few paragraphs reflecting on what was learned from your work.

Network Analysis[edit]

Either alone or in groups of up to 4, you will design and carry out a research project based on existing network data. Group projects are expected to be more ambitious than individual projects.

Basically, you will identify a network and a set of hypotheses/research questions that you would like to answer using data from that network. You will use network analysis and network visualization tools to answer those questions, and then you will report on and discuss your findings.

Finding Networks[edit]

There are a number of repositories of network datasets. Probably the simplest to use are those found in the networkdata package, which we have already been using. If you look at the homepage of the package, you can see descriptions of the datasets that are included.

Another repository I really like is the Index of Complex Networks. You can filter by type of network, size, etc. One thing to note about these networks is that you will need to load them into R (unlike networkdata). They are in different formats, and you will likely need to do some work to get them to be igraph/tidygraph networks.

Fortunately, there are a number of tools in igraph designed to help you to import different file types. This page gives a great overview.

Early in the semester you should identify the network you want to work with and make sure that 1) it has the attributes you need to answer the questions you want to answer and 2) you are able to load it into tidygraph format.

Presentation[edit]

As with the intervention assignment, I would like you to present a short (5 minute) presentation about your project toward the end of class.

Executive Report[edit]

You will prepare a ~5 page report, with the following sections:

  • Introduction
    • Give a compelling introduction to the project: why is this something worth studying? You may also want to preview some of the results of your research.
  • Network Theory
    • In your introduction, you will likely mention some network theories that helped to motivate your research. In this section, you will provide more detail about the theories, leading up to the development of 2-5 hypotheses or research questions. You should include at least three references to support your arguments.
  • Method and Data
    • You will explain how you are going to answer your research questions / hypotheses. Explain what data you will be using, and describe what analyses you will do, and why those are appropriate for answering your questions.
  • Analysis
    • For each hypothesis / research question, you will perform some sort of analysis. For some questions, this will be an analysis of network measures. For others, this might be a network visualization. Your analysis can include other types of analyses (e.g., statistical analyses) but it should include at least one network analysis and at least one visualization.
  • Discussion and Conclusion
    • Finally, write a few paragraphs 1) summarizing the findings from your analyses and 2) reflecting on how your findings fit in (or don't) with network theories you learned about in the class.



Company Pitch[edit]

Based on your work in this class, you will craft a pitch to help an organization make sense of their own social networks and craft a portfolio of your work with data drawing on the previous labs and readings to support your pitch.

  1. Select a potential employer, and research them to identify the pressing organizational issues facing them as a company.
  2. Write a document for a potential employer that
    1. Introduces you and states your professional objectives. (1 paragraph)
    2. Makes the case for the most important problems facing this organization or industry based on your research. Be sure to cite any literature that you use. (1 paragraph)
    3. Explains how a network approach and your expertise in particular can help address the sorts of problems facing their organization and industry
      • Include at least three distinctive insights (e.g., my expertise in network analysis can help you identify opinion leaders in this organization).
      • Include example visualizations of each insight from your own work (or create new visualizations).
      • Include ideas from and references to the class readings to make your case.
    4. Closes with a request for a meeting to learn more about their organization.
  3. Compile and submit your document on Brightspace

While I want everyone to write and turn in their own document, I encourage you to help each other with the visualization portions and to ask questions on the group Discord channel. I expect that the final document will be ~5 pages long (maybe longer if you want to showcase a lot of visualizations).

Sample Business Data[edit]

If you would like, you can use data from a real organization. There are some great options of network data files already in igraph format at https://github.com/schochastics/networkdata.

I created an example of how to load this data, and how to run some simple network metrics and visualizations. There's an RMarkdown file here and the HTML output here.

Examples[edit]

This project is similar to a final project that Josh Barbour assigns to his class. You can find examples of their projects on Brightspace.