YouTube COVID-19 Vaccine Misinformation on Twitter: Platform Interactions and Moderation Blind Spots (2208.13000v1)

Abstract: While most social media companies have attempted to address the challenge of COVID-19 misinformation, the success of those policies is difficult to assess, especially when focusing on individual platforms. This study explores the relationship between Twitter and YouTube in spreading COVID-19 vaccine-related misinformation through a mixed-methods approach to analyzing a collection of tweets in 2021 sharing YouTube videos where those Twitter accounts had also linked to deleted YouTube videos. Principal components, cluster and network analyses are used to group the videos and tweets into interpretable groups by shared tweet dates, terms and sharing patterns; content analysis is employed to assess the orientation of tweets and videos to COVID-19 messages. From this we observe that a preponderance of anti-vaccine messaging remains among users who previously shared suspect information, in which a dissident political framing dominates, and which suggests moderation policy inefficacy where the platforms interact.