Researchers at Lazer Lab developed a data visualization dashboard to investigate 29 million tweets related to COVID-19 collected between January 1 and September 30, 2020 from over half a million registered American voters.

This dataset can be analyzed by the top links, domains, and keywords extracted from tweets:

● Top links displays the most shared links such as a news stories

● Top domains showcases the most shared sites such as "cnn.com"

● Top keywords features the most frequently occurring words such as "pandemic"

For a more granular view, use the Choose a stateand Choose a month dropdowns to filter either at the national level or for a specific state, as well as by month. For instance, apply filters to answer the question: What were the links shared most often by Ohio residents in April?

In the Top domains section, there is the option to see who is sharing each domain according to two demographic attributes: age and political party affiliation.

In the Top keywordssection, explore keyword trends over time. Underneath Compare specific words over time, users can input one or several keywords to plot a graph showing the frequency for those keywords in the tweets dataset. As an example, the graph below visualizes the trend for "hydroxychloroquine", compared to "masks", over time: President Donald Trump promoted it in early April; in mid-May Trump disclosed that he is taking the drug; on June 15th, the FDA revokes emergency use authorization; on July 27, Twitter deleted a Trump tweet claiming its efficacy. Use the Media Cloud feature to visualize media attention over time for a specific keyword and link through to Google Trends to explore the same keyword’s attention on Google.

For additional details about the data collection, methodology, and research reports, please read the COVID-19 Fake News Sharing on Twitter report and visit www.covidstates.org.