NOTICE: The editors of PLOS ONE retracted this study, alleging that the research contained significant methodological flaws, including inadequate sampling, unbalanced search terms, and improper network analyses. The authors disagreed with the retraction and stood by their findings, despite the editors’ conclusion that the data failed to support the paper’s claim that Donald Trump was the primary driver of vaccine misinformation.
Anti-vaccination profiles on Twitter tend to be more engaged in discussions and rely on a more interconnected social network compared to their pro-vaccination counterparts, according to new research published in PLOS One. The study also indicates that former President Donald Trump was one of the key nodes in the anti-vaccination network.
“An anti-vaccine movement is currently on the rise and has gained a momentum since vaccines have become the most important tools to solve the issues caused by the ongoing pandemic,” said study author Federico Germani (@fedgermani), a researcher at the Institute of Biomedical Ethics and History of Medicine at the University of Zurich and director of Culturico.
“As WHO Director-General Dr. Tedros defined it last year, we are not only living through a pandemic, but also an infodemic, which is caused by the uncontrolled spread of misinformation and conspiracy theories online about COVID-19 and vaccines. Since social media are the platforms in which anti-vaccine supporters gather, we were interested in understanding the behavioral dynamics that govern anti-vaccine information flow on Twitter, hoping to identify viable strategies to limit the circulation of falsehoods about vaccines.”
In the study, which included about 150 Twitter users, the researchers used the #vaccineswork hashtag to identify pro-vaccination individuals on Twitter, while the hashtags #vaccineskill and #vaccinesharm were used to identify anti-vaccination individuals. A separate control group was identified by searching for hashtags generated by a random word generator. All of the profiles were manually screened to ensure the users’ opinions reflected their given group.
Germani and his colleague, Nikola Biller-Andorno, found that the anti-vaccination group was the most active on Twitter, with 536 tweets, replies and retweets per month on average between September and December 2020. In contrast, the control group had 277 tweets, replies and retweets per month on average and the pro-vaccination group had only 144 Twitter actions per month on average.
The pro-vaccination profiles were more likely to generate new content, but the anti-vaccination profiles replied 13-times more than control and pro-vaccination profiles.
The anti-vaccination group also scored the highest number of retweets per tweet, and shared more emotional content and conspiracy theories compared to the other groups. There were also differences observed in the most commonly used words in tweets. The anti-vaccination group tended to use words such as “President,” “God,” “People,” and “Masks,” while the pro-vaccination group tended to use words such as “Help,” “Health,” “Thanks,” or “Research.”
“Our study suggests that social media polarization is a big issue, as people exposed to various conspiracy theories tend to also be exposed to anti-vaccine discourse on Twitter (and vice versa). We therefore think social media giants should play an active role to prevent polarization, including in the context of vaccines. Generally, if people are exposed to different views, they tend to better evaluate sources of information and to consider the possibility they may be wrong,” Germani told PsyPost.
By using Cytoscape software to analyze the social networks on Twitter, the researchers found that the anti-vaccination network was more interconnected compared to the pro-vaccination network. The World Health Organization was the only large influencer in the pro-vaccination network. But there were 14 large influencers in the anti-vaccination network.
“We show that the success of the anti-vaccination movement online is likely based on common beliefs and interests, through which users establish a well-connected community and constitute an echo chamber for contents generated by a smaller fraction of profiles. We define these latter users as anti-vaccination influencers. We identify former US President Donald Trump as the main influencer in the anti-vaccination web,” the researchers wrote in their study.
Despite not making overt anti-vaccination statements as president, Trump had published dozens of tweets linking vaccinations with autism in the past, such as one in 2014 that read: “Healthy young child goes to doctor, gets pumped with massive shot of many vaccines, doesn’t feel good and changes – AUTISM. Many such cases!” Research has found that Americans who voted for Trump in 2016 were particularly prone to anti-vaccination attitudes and that these attitudes were exacerbated by his tweets.
“Besides Trump, we identify his son Donald Trump Jr, Charlie Kirk, a popular evangelical Christian and Republican activist who supported Trump’s presidency, James Wood, a popular actor and producer who is also a strong supporter of Trump–to be among the largest influencers in the anti-vaccination network,” Germani and his colleagues wrote.
The researchers hope their findings can help to combat misinformation on social media.
“The general public should be concerned with increasing vaccine hesitancy, as this may ultimately impact everybody’s life. Vaccine coverage is of utmost importance, and even more now to overcome the social and economic issues imposed on us by COVID-19. Our study may help people to understand how not to fall pray of the anti-vaccine discourse on Twitter,” Germani told PsyPost.
“Our study proposed strategies to halt the ongoing anti-vaccine infodemic. These approaches are oriented towards reducing or blocking the circulation of incorrect anti-vaccine information. However, our study does not highlight long-term strategies that are needed to prevent the anti-vaccine discourse to find a fertile ground among the general public. For this reason, we are planning to study whether and which critical thinking skills could aid people to understand how to access information online, providing them with learning tools that could help them to spot scientific misinformation online.”
The study, “The anti-vaccination infodemic on social media: A behavioral analysis“, was published March 3, 2021.