New article - Facebook post data - A primer for educational research

Joshua Rosenberg


With my colleagues Conrad Borchers and Rita Swartzengruber, I published an article in Educational Technology Research and Development on using public Facebook post data for research. The article is titled Facebook post data: a primer for educational research. Here’s its abstract:

Facebook is widely used and researched. However, though the data generated by educational technology tools and social media platforms other than Facebook have been used for research purposes, very little research has used Facebook posts as a data source—with most studies relying on self-report studies. While it has historically been impractical (or impossible) to use Facebook as a data source, the CrowdTangle platform allows academic researchers to freely access the massive collection of posts on public Facebook pages and groups. In this paper, we first outline how interactions and textual features in these public Facebook data in concert with established methods from educational data mining and learning analytics can be used to scrutinize educational discourse and knowledge sharing at scale. We then provide a primer that offers considerations for researchers before collecting these data (i.e., conducting research ethically and framing the study). The tutorial also covers matters directly pertaining to using CrowdTangle: accessing the CrowdTangle platform, uploading or identifying pages (or groups), and downloading historical data and it includes code using the statistical software and programming language R. We conclude with ideas for future directions for using Facebook posts as data with a focus on how educational researchers can leverage the scale of the available data and the time periods for which data is available to study educational affairs (i.e., issues or topics) and individuals (i.e., people or organizations) and to scrutinize how Facebook itself is used.

It’s presently available open-access here, but I think that will change when the article is included in an issue. A pre-print is here. The R code we used to analyze the data is available here

This paper stemmed, originally, from trying to understand schools’ and districts’ responses to COVID; when my colleagues and I realized that school websites had limited up-to-date information, but that these institutions were highly active on Facebook, we turned to trying to access public Facebook post data at scale.

We–Conrad, Macy Burchfield, Sondra Stegenga, Daniel Anderson, and Christian Fischer–did this through a tool called CrowdTangle that is open to academics, soon realizing that there were, we thought, notable violations of students’ privacy. This work was covered by a range of news outlets, including Wired and EdWeek.

When stepping back from the specific use cases we envisioned for Facebook post data, we thought it might be valuable to others to consider how they could use Facebook data for research related to teaching, learning, and educational systems, thus the primer we just published.

Especially as Twitter/X data becomes harder (maybe not possible) to access, I hope that Facebook/Crowdtangle remains a possibly informative data source for educational researchers. If you are an educational researcher reading this and you want to use CrowdTangle and don’t have or can’t gain access, let me know; we’ll see if I can help!