Twitter Dataset: Replies to President Trump’s Tweet about contracting the Novel Coronavirus

On October 2nd, 2020, U.S. President Donald Trump tweeted that he and the First Lady of the United States (FLOTUS), Melania Trump, both tested positive for COVID-19 (https://twitter.com/realDonaldTrump/status/1311892190680014849). Within seconds, his tweet received thousands of replies. The dataset consists of 298,172 replies to Donald Trump’s tweet announcing his COVID-19 diagnosis, posted on October 2nd between 6am and 12:30pm (ET). It contains tweet ids, the toxicity scores (as calculated by Google's Perspective API via https://Communalytic.com) and tweet availability status values (via twarc library). Following Twitter’s API policy, we stripped metadata associated with each tweet. So, if you’d like to examine potential relationships between other metadata elements, you would need to recollect original tweets using tools like DocNow’s Hydrator first. The only downside of this approach is that tweets that have been blocked or deleted will not be recollected. To help you get started, we also shared our Exploratory Data Analysis (EDA) Python script at https://github.com/RUSocialMediaLab/toxicityanalysis