Researching the Effects of Online Misinformation and Realities of the Coronavirus

Assistant Professor Cody Buntain, NJIT's expert on the science of understanding social media, is turning his trained eyes to the connections between online misinformation and public health crises such as the COVID-19 pandemic.

Using software that he co-developed in 2017 with University of Waterloo Professor Jimmy Lin, which identifies high-impact moments on Twitter, Buntain will compare relevant tweets about the coronavirus to actions among the people of countries including Brazil, Italy, New Zealand and the United Kingdom.

His research is supported by a six-month, $138,000 grant from the U.S. Department of Energy. The department's Oak Ridge National Laboratory and the Department of Defense's National Geospatial-Intelligence Agency are also research partners because of their expertise in large-scale data analytics and missions to protect American interests worldwide.

"We started discussions about this project back in March as part of a rapid response to the COVID-19 outbreak. What we wanted to understand was how can we combine our understanding of what happens on social media with public health outcomes," Buntain said, referring to situations such as people discussing or fiercely arguing the pros and cons of wearing masks, quarantining or taking vaccines.

"The expectation is there is some correlation here," Buntain observed. Typically, one would expect that the more misinformation gets published online, the worse it is for public health outcomes, he said.

The study is just getting started, so the exact correlations won't be known until next spring. Buntain said he expects interesting results because online spaces tend to amplify extreme views that are not representative of general populations. They're not examining U.S. trends, partially because American users are too often influenced by individual accounts that have outsized influence. If one takes an optimistic view, "It could be the case that the public is more resilient to this kind of thing than we expect," he noted.

Outside of their research grant, Buntain was asked about the pandemic's social media influence for the upcoming presidential election. "I might almost go the other way — how is it that the election affects conversations about COVID? This makes for really strange behaviors," he said.

"How it dominates the conversation around the election is definitely an open question. … You can imagine when people start talking about trying to go vote in person, what are the COVID-19 health-related risks of that? That's potentially something that we could track," he said.

"One of the really interesting questions here that we are trying to answer is about how people who are in the same country, but in different regions of the same country, can have different views and outcomes of the conversations they are having. If we were looking at the U.S., you might consider how people in Florida are responding to this versus people in New York or New Jersey."