[Paper Review] One Year of COVID-19 Vaccine Misinformation on Twitter: Longitudinal Study

Francesco Pierri, Matthew DeVerna+4 more|arXiv (Cornell University)|Sep 4, 2022

Misinformation and Its Impacts4 citations

TL;DR

This longitudinal study analyzes COVID-19 vaccine misinformation on Twitter from January to December 2021, using automated detection and network analysis to track low-credibility content and sharing patterns. It reveals that a small group of ~800 verified 'superspreaders'—notably Robert Kennedy Jr.—accounted for 35% of all misinformation reshare volume, with suspicious YouTube links and bot accounts amplifying harmful content, suggesting financial incentives drive dissemination despite deplatforming efforts.

ABSTRACT

Vaccinations play a critical role in mitigating the impact of COVID-19 and other diseases. This study explores COVID-19 vaccine misinformation circulating on Twitter during 2021, when vaccines were being released to the public in an effort to mitigate the global pandemic. Our findings show a low prevalence of low-credibility information compared to mainstream news. However, most popular low-credibility sources had larger reshare volumes than authoritative sources such as the CDC and WHO. We observed an increasing trend in the prevalence of low-credibility news relative to mainstream news about vaccines. We also observed a considerable amount of suspicious YouTube videos shared on Twitter. Tweets by a small group of about 800 "superspreaders" verified by Twitter accounted for approximately 35% of all reshares of misinformation on the average day, with the top superspreader (@RobertKennedyJr) responsible for over 13% of retweets. Low-credibility news and suspicious YouTube videos were more likely to be shared by automated accounts. Our findings are consistent with the hypothesis that superspreaders are driven by financial incentives that allow them to profit from health misinformation. Despite high-profile cases of deplatformed misinformation superspreaders, our results show that in 2021 a few individuals still played an outsize role in the spread of low-credibility vaccine content. Social media policies should consider revoking the verified status of repeat-spreaders of harmful content, especially during public health crises.

Motivation & Objective

• To understand the dynamics of low-credibility vaccine misinformation on Twitter during the 2021 pandemic.
• To identify key actors—particularly verified users—responsible for amplifying misinformation.
• To examine the role of automated accounts and suspicious YouTube links in spreading misinformation.
• To assess the relative reach of low-credibility sources versus authoritative health organizations like the CDC and WHO.
• To evaluate the effectiveness of social media verification and platform policies in curbing harmful content dissemination.

Proposed method

• Collected and analyzed 1.2 million tweets related to COVID-19 vaccines from January to December 2021 using Twitter’s API.
• Classified sources as low-credibility or mainstream news based on fact-checking databases and media credibility assessments.
• Identified 'superspreaders' as verified users who generated disproportionately high reshare volumes of misinformation.
• Tracked and analyzed the sharing patterns of suspicious YouTube videos embedded in tweets.
• Used bot detection heuristics to assess the prevalence of automated accounts in resharing low-credibility content.
• Applied longitudinal analysis to measure trends in misinformation prevalence relative to mainstream news over time.

Experimental results

Research questions

• RQ1 What was the prevalence of low-credibility vaccine misinformation on Twitter relative to mainstream news during 2021?
• RQ2 Which verified users were responsible for the highest volumes of resharing misinformation?
• RQ3 How did the reach of low-credibility sources compare to authoritative sources like the CDC and WHO?
• RQ4 To what extent were suspicious YouTube videos and automated accounts involved in spreading misinformation?
• RQ5 What incentives might drive the behavior of high-volume misinformation spreaders?

Key findings

• Low-credibility vaccine misinformation was less prevalent overall than mainstream news, but it achieved significantly higher reshare volumes.
• The top 800 verified 'superspreaders' accounted for approximately 35% of all daily reshare volume of misinformation.
• Robert Kennedy Jr. alone was responsible for over 13% of all retweets of misinformation on average per day.
• Low-credibility sources and suspicious YouTube links were more frequently shared by automated accounts than by human users.
• The prevalence of low-credibility vaccine content relative to mainstream news showed a statistically significant increasing trend over the course of 2021.
• Findings support the hypothesis that financial incentives drive the behavior of superspreaders, despite deplatforming of some high-profile accounts.