In a study published in the Journal of Medical Internet Research, scientists used machine learning techniques to analyze hundreds of thousands of Twitter coronavirus-related messages from Republican and Democratic lawmakers. The researchers found that Democratic tweets tended to focus on racial disparities in care and the Trump administration, while Republican tweets tended to focus on vaccine development and China.
“In our prior work using Twitter, we found that expressions of loneliness, depression, anxiety, and general sentiment changed significantly across counties in the United States after the declaration of national emergency in March 2020,” said study author Sharath Chandra Guntuku, an assistant professor at the University of Pennsylvania.
“In this study, we wanted to see how U.S. legislators were taking to Twitter to share pertinent information about COVID-19, specifically around healthcare, and how they frame it.”
For their study, the researchers analyzed more than 300,000 tweets posted by 4,224 state lawmakers from January 1 to October 10, 2020. Guntuku and his colleagues were able to categorize various coronavirus-related topics that the lawmakers tweeted about, as well as whether the tweets contained positive or negative sentiments.
“There are several differences and also a few commonalities in the themes (and associated sentiment) that U.S. legislators across party affiliations shared on Twitter about COVID-19. For example, Democrats focused on racial disparities in care and protections for essential workers whereas Republicans focused on vaccine development and economic aid,” Guntuku told PsyPost.
Democrats were more likely than Republicans to use words such as “black,” “housing,” and “workers.” Republicans, on the other hand, were more likely than Democrats to use words such as “CCP” (the abbreviation for the Chinese Communist Party) “trial,” and “food assistance program.” Republicans also used more positive sentiment in their tweets, which increased over time.
​”In today’s digital era, we are seeing more and more that individuals go to social media for their news and understanding how legislators use these platforms to communicate is important. Here we find key differences in Republicans and Democrats messaging specifically around COVID-19,” added co-author Anish Agarwal.
Future research could examine how legislators are using language on other social media websites, such as Facebook.
“We only used data from Twitter; legislators might be using different platforms to share with different audiences, which could vary by demographics and affiliation,” Guntuku said.
The study, “Partisan Differences in Twitter Language Among US Legislators During the COVID-19 Pandemic: Cross-sectional Study“, was authored by Sharath Chandra Guntuku, Jonathan Purtle, Zachary F. Meisel, Raina M Merchant, and Anish Agarwal.
