This project seeks to understand the nature of misinformation on social media and its impacts on offline political behavior. Using the tools of computational social science, this project identifies, tracks, and detects misinformation related to the Covid-19 pandemic on Twitter. Examining how misinformation spreads relative to true information about the pandemic allows the project to understand why misinformation spreads, whether it can be detected in real time, and what its effects are on offline political behavior.
This project employed three undergraduate students for the summer of 2020 and fall semester 2020. These students were Gaurav Sett, Abdulwahab (Sam) Jarkas, and Lindsay Schurtz. The students contributed to building code to scrape and analyze tweets taken from the Twitter social media platform. Tweets were analyzed with respect to content that was deemed to be misinformation related to the Covid-19 pandemic. The students built algorithms to identify and track the spread of misinformation about Covid-19 from February 2020-June 2020. They were able to detect a sizable spike in Covid related misinformation in early 2020, which quickly subsided where it returned to a baseline of just a few percentage points of all tweets collected by March 2020.
The project has built tools and identified methods to detect misinformation in social media data streams. Further research will correlated the spread of this misinformation with public attitudes and violent events that occur offline.
We have not yet published any research related to this project, but I have successfully received further DILAC and GTRI HIVES IRAD funding to continue the nature of this project by investigating the link between misinformation spread on social media and radicalization by social media users against the democratic process in the United States. This includes events like violent protests against public health measures throughout 2020 to the Capitol insurrection in 2021.