Imprimir Republish


On the trail of misinformation

Users aligned with the conservative right were more engaged in disseminating false news and inaccurate information on Twitter, according to a survey by FGV

Alexandre Affonso

People aligned with the conservative right were the most engaged in disseminating false news and inaccurate information about Covid-19 on Twitter, according to a survey conducted during the height of the pandemic by researchers at the Getulio Vargas Foundation (FGV), in Rio de Janeiro. They evaluated 3.3 million posts on the social network between January and May 2021. Based on their analysis of the links shared and the individuals who shared them, they defined four profile groups (see infographic). One of the groups, made up of politicians, bloggers, and activists from the socially conservative right—equivalent to 21.5% of the survey sample—accounted for almost half of the interactions analyzed. Most of their posts advocated the use of drugs that have no effect against the 2019 coronavirus (Sars-CoV-2).

The posts by users in this group generated the most comments and shares. They also circulated for longer periods on the social network, and almost always linked to information produced by websites with journalistic facades, which presented biased content designed to attribute some scientific credibility to their assertions and positions. “These publications had links to anonymous pages devoted to the dissemination of false or intentionally distorted information with the purpose of trying to legitimize the effectiveness of drugs that are part of what has come to be called ‘early treatment’,” explains sociologist Victor Piaia, from the Public Policy Analysis Board (DAPP) at FGV, one of the survey authors. The work was developed within the scope of the project called Digital Democracy: Digitalization and the Public Sphere in Brazil, which has the support of the German embassy in Brasília and the Ministry of Foreign Affairs of Germany.

These links are often attached to exaggerated headlines. The links that circulated the longest on Twitter during the survey period linked to the website, which, in March, released a study “confirming” the efficacy of the antiparasitic medicine ivermectin against Covid-19. The news gained momentum on the social network, generating intense user engagement, which caught the attention of fact-checking agencies. One of them showed that the study was, in fact, a synthesis of results from other research carried out by the website itself. The survey had several methodological flaws. It grouped non-comparable studies, many published in preprint format (lacking peer review), while ignoring the results of robust studies that did not identify significant effects for the use of ivermectin against Sars-CoV-2. “The account responsible for launching the content was suspended, but the link to the website continued to be shared, circulating on Twitter for 159 days,” Piaia says.

Another link that circulated widely among the members of this grouping refers to a study published in 2005 in Virology Journal. According to the accompanying headline, the work proves the effectiveness of the antimalarial drug hydroxychloroquine in the treatment of Covid-19. However, the fact-checking team from the Reuters news agency found that the experiment had been carried out on animals, not humans, and that it was focused on Sars-CoV, which causes Severe Acute Respiratory Syndrome (SARS), a different disease than Covid-19, which is caused by the Sars-CoV-2 virus. Piaia points out that users in this grouping, more than users in the other groups, have difficulty understanding how science works. “At the same time,” he says, “we noted that these users value the credibility of science and use it to try to support their positions and legitimize their opinions.” Often, the posts also make use of statements from “international experts” to support the effectiveness of the remedies, as if the fact that they are from abroad gives them greater reliability.

For Raquel Recuero, a researcher at the Media, Discourse, and Social Network Research Lab at the Federal University of Rio Grande do Sul (MIDIARS-UFRGS), the results of the FGV study reinforce the conclusions of other studies, which show that specific discursive strategies, such as scientific authority, are being widely used on social media to try to legitimize the dissemination of false information. “This is a common strategy being used a lot during the pandemic as a way to counter scientists, using pseudoscientific arguments that almost always assert a differing interpretation,” she says. These studies also found that users who disseminate misinformation tend to be more engaged, sharing and commenting on the content much more frequently. “The analyses by these studies indicate that the probability of a link advocating the use of hydroxychloroquine against Covid-19 being shared on Twitter is almost three times greater than that of a link with content that challenges that premise.”

Alexandre Affonso

The persistence of defending hydroxychloroquine and ivermectin is a Brazilian phenomenon. In other countries these drugs stopped attracting attention when their ineffectiveness was demonstrated, and the misinformation on social networks in these countries changed its focus primarily to the supposed risks of vaccines. In Portugal, the website SciMed – Ciência Baseada na Evidência (Science Based on Evidence) stood out by denouncing and combating misinformation propagated by health professionals associated with alternative therapies. One group called Doctors for Truth claimed, for example, that tests to detect Covid-19 were unreliable and that the epidemic was being exaggerated. “Most of them are people connected to pseudoscience,” said doctor João Júlio Cerqueira, the site’s creator, in an interview on the Polígrafo website. The Portuguese Medical Association has opened disciplinary proceedings against some of these professionals.

One of the most striking dynamics of publishing and sharing on Twitter, according to the FGV study, is the fact that it is restricted to specific audiences. This endogenous dynamic can be identified in the community that defended the effectiveness of so-called early treatment, although it also manifests in the grouping of users located on the left of the political spectrum, composed of politicians, celebrities, and social activists who oppose the federal government. “The problem is that links from the traditional media and fact-checking agencies can only penetrate groups in which disinformation circulates when the published content is aligned with the opinions and ideologies of those users,” says Felipe Soares, a researcher from MIDIARS-UFRGS. “Most of the time, the links produced in each bubble, whether right or left, tend to circulate only among the users of that group.” He believes this phenomenon results from the fact that the pandemic in Brazil, as in other countries, is perceived by various groups as a political-ideological issue and not as a public health issue. “The discourse around treatment strategies and actions to contain the virus is polarized and, therefore, tends to be confused with being a matter of political party affiliation,” emphasizes Soares.

The discussion about disinformation in social networks becomes increasingly complex the more its consequences are studied. One of the most problematic points concerns the boundary between what constitutes false information and imprecise information. There are cases where a news item has some truth to it but is presented out of context or accompanied by extreme interpretations. “Many users do not perceive these differences,” says Piaia. “Often, people share posts simply because they reinforce their point of view.” At the same time, others spread disinformation to gain economic advantage, influence, and prestige. There have been recent cases of digital influencers sharing false information to make money. As reported by Agência Pública (a Brazilian investigative and independent journalism agency) in March, the Brazilian Ministry of Health and the Secretariat of Communication paid more than R$1.3 million to internet personalities to publicize campaigns about Covid-19 on their social networks, R$23,000 of which went to those who promoted “early treatment.”

The researchers themselves find it difficult to pinpoint who is propagating the fake news or even identify their motives. This is because the automated tools they use to collect their sample data are based on keyword identification. The boundary between profiles of real people and remote-controlled profiles, known as bots, is also blurry. Piaia says there now exist many real profiles on social networks that behave like bots, and bots that are increasingly good at imitating real users. “Detection methodologies need constant improvement,” he adds. “But it is possible to verify that much of the disinformation circulating on social networks is born in nuclei made up of more radical Twitter profiles, then reverberates through the profiles of common users, who often aren’t radical but see some meaning in this content they want to share.”

Another challenge is to understand why the conservative right is almost always the engine of disinformation propagation and why fake news exerts a tighter grip on people aligned with this political spectrum. For some experts, part of the answer lies in a general suspicion these individuals have of solid sources of information such as traditional news organizations and scientists.

Dayane Machado, a doctoral student at the Department of Science and Technology Policy at the University of Campinas (UNICAMP), observes that media manipulation campaigns are usually organized by small groups that produce materials with a specific target in mind, who then try to attract the attention of celebrities and politicians to use to amplify their message. “One of the ways to attract the attention of these celebrity profiles that have lots of followers is to give the misinformation a scientific façade, so that non-specialists believe it and pass on the material as if it were legitimate,” the researcher explains.

The traditional news media can itself be used to spread disinformation when headlines for news stories are poorly written

One of the difficulties of containing the spread of disinformation on social media is to increase the reach of quality content, and penetrate bubbles predominantly fueled by fake news. The problem, explains Machado, is that the actors responsible for producing this type of content—journalists, scientists, educators, etc.—work at a different pace and in an environment mediated by the cautious analysis of data, while the producers of fake news, who have no commitment to the truth, can react much more quickly and aggressively. A study published in the journal Science in 2018 by researchers at the Massachusetts Institute of Technology (MIT), in the United States, reinforces this idea. They analyzed 126,000 news stories, both false and true, shared by approximately 3 million people on Twitter from 2006 to 2017. They found that false news items are produced more quickly and therefore always appear to be more recent, and that users on the social network are more likely to share newer information. “Sham news stories travel further and faster than news that went through the verification process, and the engagement around them was greater when they dealt with political issues,” the researchers wrote.

Traditional news outlets can also be used to spread disinformation when the headlines for stories are poorly written and allow for widely differing interpretations, according to a study published in March in M/C Journal. An analysis of 20 internet addresses for the Brazilian pandemic news stories that were most shared within 1,632 Facebook groups during 2020 showed that for 43.8% of them, users were only interested in content whose titles could reinforce some type of distorted information. In these groups, 86.2% of the messages written by users about the stories reproduced some type of misinformation that minimized the severity of the pandemic, containment measures, and vaccines.

Educating against fake news

Media education can be a tool in combating misinformation, leading to public understanding of the news production process, how information sources are chosen, and how media outlets function. “With the increase in the propagation of fake news, the debate about the importance of media education has increased,” assesses journalist Ivan Paganotti, coordinator of the Verification, Education, Communication, Algorithms, and Regulation research group, at the Methodist University of São Paulo. “The more people know how the news media works, the more they will be able to consume news critically.”

Paganotti is a current affairs teacher at the São Paulo university. With his students, he discusses the importance of selecting verifiable sources of information, inviting them to reflect on the news and to look for information in more than one news outlet, with complementary perspectives. “I always tell them to ask themselves how the content got to them, which source originated the information and if that source has credibility, and if it usually publishes corrections when it makes an error. Does the piece leave space for conflicting information; for example, if a person was accused of something, was their viewpoint heard? Is the publication date recent? These elements help with deciding if the content is trustworthy,” he explains.

In 2018, Paganotti and fellow researchers created, with Facebook support, the free online course Vaza, Falsiane. In addition to explaining the pitfalls of misinformation, in 2021 the course added a special module on misinformation and the pandemic. Their experience with the project was detailed in an article published in 2021 in the journal Intexto, from UFRGS. With profile pages on the major social networks, the course publishes posts with alerts and humorous tips. Paganotti notes, “We can’t fight disinformation on our website alone, we need to take the battle to the territory where it circulates.”

Scientific article
Vosoughi S., Roy D. & Aral S. The spread of true and false news online. Science. Mar. 2018.