Research Guides: Social Media Research Guide: Sources of Social Media Data

Apply for access to CrowdTangle for Facebook and Instagram

Apply for access to CrowdTangle
CrowdTangle tracks data from public content across Facebook Pages and Groups, as well as Verified Profiles and public Instagram accounts.

Faculty, PhD students and post-docs doing research in the following areas may apply:

Misinformation
Elections
COVID-19
Racial justice
Well-being

What Does Twitter Make Available?

For getting the most out of Twitter's search interface

Twitter advanced search (help available here)

Twitter API

Update 2/13/23 Twitter announced the Academic Research API will be replaced with new "low-cost, basic access that offers a low level of API usage, and full access to Ads API for a $100 monthly fee". More information will be available on the website

Twitter Curated Datasets for Researchers

As of 2/13/23 Twitter says it is exploring "free, purpose-built, low-code datasets" for research. More information is available here and sign up to be notified when it's available

Curated Twitter Collections for Research

SOMAR - Social Media Archive at ICPSR
SOMAR's mission is to provide a centralized and reliable repository for social media research data. The archive contains a range of data collected from large-scale social media platforms such as Twitter, Facebook, Instagram, and Reddit. The data in the archive takes two forms. They can be public, i.e., available for immediate download, and/or restricted, i.e., available within our secure data enclave after receiving approval following a submitted restricted data application.
TweetSets (George Washington University)
Collection of large Twitter datasets (tweet IDs) on a variety of newsworthy topics, including a number of political and election-related topics (primarily US) back to 2016. To use them, you have to rehydrate the tweets using a tool like DocNow’s Hydrator: https://github.com/DocNow/hydrator.
DocNow Tweet Catalog
The DocNow Catalog is a collectively curated listing of Twitter datasets. Public datasets are shared as Tweet IDs, which can be hydrated back into full datasets using the DocNow Hydrator desktop application.

Brandwatch (U-M Access Only) for Twitter, Reddit, Tumblr

Brandwatch replaces Meltwater for social media searching and analytics

The Library has 10 queries/projects that we can make available to our researchers in Brandwatch, a social media listening tool. Brandwatch has full firehose data for Twitter, Reddit and Tumblr back to 2010. Brandwatch does not have Facebook or Instagram data.

Fill out this form to request access to a query. Brandwatch can be used for academic research only.

Your access to Brandwatch will be limited in time and then your query will be deleted to make room for the next interested researcher. Brandwatch sets strict limits on downloads. Brandwatch is not a good source for exporting large amounts of tweets, the Twitter API or Twitter Decahose are better options.

Brandwatch can be used by researchers from the Ann Arbor, Flint and Dearborn campuses.

If you're affiliated with Ross you should access Brandwatch through the account managed by the Kresge Library

Twitter Decahose (U-M Access Only)

University of Michigan researchers can access a compilation of tweets known as the “Decahose” (a 10% sample of all tweets) without charge. MIDAS, CSCAR and ARC-TS together manage and support the use of this data repository, including the historical archive of Decahose tweets and ongoing collection from the Decahose.

U-M researchers can use this set of data for five areas of research: information diffusion; Natural Language Processing; network analysis; behavior analysis; Sociolinguistics.

See the MIDAS website for more information and to inquire about access