The public debate on social distancing in the context of COVID-19: analysis of an communities of attention on X (Twitter)

ABSTRACT

Introduction:  During the COVID-19 pandemic, scientific research began to be widely shared on social media, with particular emphasis on the social network X. Altmetrics helps to understand the impact of these studies and the profile of the communities that interacted with them in this environment.

Objective:  This case study aims to map the communities of attention network that interacted with the most shared COVID-19 article on X in Brazil between 2020 and 2022.

Methodology:  The article "Social Distancing Alters the Clinical Course of COVID-19 in Young Adults: A Comparative Cohort Study" was selected from the Dimensions database. Altmetric.com was used to track mentions on X, excluding retweets without comments. The data were analyzed using the VOSviewer software.

Results:  A total of 483 mentions on X were analyzed, with 7.03% being original tweets and 92.96% being commented retweets. The communities of attention network included 470 users, with 12 verified accounts. Most profiles were individual, with a significant presence of academic and scientific profiles, who showed concern for sharing science-based information.

Conclusion:  The communities of attention network on X proved to be committed to disseminating legitimate scientific information about health and protection against COVID-19.

KEYWORDS:
COVID-19; Metric studies of information; Scientific communication; Communities

RESUMO

Introdução:  Durante a pandemia da COVID-19, pesquisas científicas começaram a ser amplamente compartilhadas nas mídias sociais, com destaque para a rede social X. A altmetria ajuda a entender o impacto dessas pesquisas e o perfil das comunidades que interagiram com elas nesse ambiente.

Objetivo:  Este estudo de caso visa mapear a rede de comunidade de atenção que interagiu com o artigo sobre COVID-19 mais compartilhado no X no Brasil entre 2020 e 2022.

Metodologia:  O artigo "Social Distancing Alters the Clinical Course of COVID-19 in Young Adults: A Comparative Cohort Study" foi selecionado na base de dados Dimensions. Utilizou-se o Altmetric.com para rastrear menções no X, com exclusão de retweets sem comentários. Os dados foram analisados com o software VOSviewer.

Resultados:  Foram analisadas 483 menções no X, sendo 7,03% tweets originais e 92,96% retweets comentados. A comunidade de atenção incluiu 470 usuários, com 12 contas verificadas. A maioria dos perfis eram individuais, com destaque para perfis acadêmicos e científicos, que demonstraram preocupação em compartilhar informações baseadas na ciência.

Conclusão:  A comunidade de atenção no X mostrou-se comprometida em disseminar informações científicas legítimas sobre saúde e proteção contra a COVID-19.

PALAVRAS-CHAVE:
COVID-19; Estudos métricos da informação; Comunicação Científica; Comunidades

1 INTRODUCTION

In the face of the COVID-19 pandemic, information on the topic has been widely disseminated, impacting science, academia and the daily lives of people around the world. Scientific research is disseminated to the public through social media platforms such as X (formerly Twitter)¹, blogs, news portals, and other web sources, facilitating interactivity between individuals and research findings, and highlighting the public debate on health information.

Amaral (2020, p. 238) highlights the importance of access to these media, which "[...] allow citizens to give visibility to issues that directly affect their daily lives", such as the pandemic. As a result, erroneous data, scientific misinformation and fake news have also gained a lot of ground, as it has become easier for any individual to spread and access information, whether real or not, beyond the academic sphere; the debate on issues related to COVID-19, such as its effects, statistics, treatments, containment measures, spread and social isolation, has become urgent (Leal; Ribeiro, 2020).

This disease has been of great interest to the scientific community, with a large amount of literature being continuously made available (Shamsi; Lund; Seyyedhosseini, 2022). This new reality has led to an increase in the use of social media as an alternative environment for debates on topics related to the virus, including protection and hygiene measures (Yousefifinaghania et al., 2021). This has allowed different types of analysis of scientific production to stand out, such as altmetrics and article-level metrics.

According to Gouveia (2013, p. 224), "altmetric data on access to articles and comments on them can be used to monitor the interest and relevance of published content over time. They allow us to see the demand for this work in a virtual environment, where the more a document is shared on web sources, the more interesting and important it becomes to the academic community, which can be an indicator of the impact of the research (Ortega, 2016).

Such studies should have a quantitative as well as a qualitative and social perspective, helping to reveal aspects of academic and general public interest in the research results disseminated on social networks. They are even more necessary in times of social anomalies, such as the coronavirus pandemic, in which "[...] as important as monitoring production, it is equally fundamental to monitor the dissemination and social circulation of research related to the topic, to understand who is talking about it, as well as what is being said" (Araújo et al., 2023, p. 2).

Thus, the relevance of this research is justified as a contribution to the field of information science, both because of the growing use of social media for the dissemination and sharing of scientific research in the context of scientific communication (Peixoto, Araújo, 2022; Silva; Cendón, 2022), and as a methodological contribution to the study of altmetrics, which analyzes the engagement that scientific research has shown in environments beyond the academic (Holmberg et al., 2014). As a contribution at the social level, the aim is to highlight the behavior of society through online community of attention networks in relation to the sharing of scientific research on COVID-19 in an environment that aims to socialize public debates and generate networks of relationships, such as X.

2 ALTMETRICS AND THE RELATIONSHIP BETWEEN ARTICLE-LEVEL METRICS AND SOCIAL MEDIA

Altmetrics collects statistics on scholarly publications circulating on the Internet, such as data on usage, storage, sharing, mentions of research, and allows this data to be enriched with demographic and profile information (Adie; Roe, 2013; Das; Mishra, 2014). It also makes it possible to measure the acceptance of and interest in productions by both the academic and general public (Haustein et al., 2014; Thelwall; Wilson, 2015).

The analysis is carried out using specific virtual tools, such as the Altmetric.com² system, which, by retrieving a wide range of results, provides statistical reports that consider the type of engagement that the scholarly production has received: by the academic or the general public, and the type and importance of the disseminating source (Vanti; Sanz-Casado, 2016). To achieve this, publications need to have some kind of unique digital identifier, such as the Digital Object Identifier (DOI).

According to Adie and Roe (2013), the mission of altmetrics is to facilitate article-level metrics. The latter are considered qualitative and quantitative indicators that show the impact an individual article has had. They measure information such as the number of readers the individual article has reached and it is targeting by collecting data from different sources such as social media, news portals and blogs (Adie; Roe, 2013; Das; Mishra, 2014; Trajkovski, 2016).

The social network X is considered the largest source of social media in terms of academic conversations in the world (Liu; Adie, 2013; Mendes; Maricato, 2020). Although it does not provide data on the number of users, it is estimated that 20 million people in Brazil use it mainly to share information (Hising, 2024).

Its mentions are used as a source of study, as "[...] more and more researchers are using this type of tool to publish their research or to interact with other researchers and follow the indications of references of interest to the field in which they work" (Gouveia, 2013, p. 222). Its users share information and chat in real time and can mention research soon after it is published, increasing the number of downloads and citations (Bik; Goldstein, 2013; Bornmann; Haunschild, 2018).

The use of these sources has made the process of scholarly communication more common, as they make it easier for a published document to reach its potential audience and measure the extent to which it circulates and is disseminated through communication channels, which can be through responses to messages, comments, discussions, mentions, and shares (Orduña-Malea; Martín-Martín; Delgado-López-Cózar, 2016).

According to Liu and Adie (2013), article-level metrics are excellent tools for identifying online communities that share scientific productions. Through discussions on blogs and conversations on social media, they can serve as indicators of the influence of publications, allowing a deeper understanding of the usage patterns of scholarly communication tools.

3 ONLINE COMMUNITIES OF ATTENTION

One of the ways to understand and get to know online communities of attention is through altmetrics. Araújo (2020) discusses the concern to contribute to analyses and discussions that show where and how articles are used in the online environment by different communities. Thus, they can be useful for validating content of interest to different fields, as their members have specific needs and require specialized interactions (Bik; Goldstein, 2013).

Liu and Adie (2013) explain that although publishing research in the form of journal articles is a purely academic activity, discussing and sharing these articles through online platforms such as X can also be an action carried out by non-specialists, in which it is possible to measure the impact or attention of specific groups of users outside of academia. This is the most interesting way to apply altmetrics, since it uses usage data from these sources to classify communities (Bornmann; Haunschild, 2018).

To understand the community of attention in social media, especially in X, certain types of information can be gathered from user profiles, such as their biography, in which they describe themselves and their various interests. It is also possible to analyze tweets (posts on the X social network that can contain text and multimedia) as an additional source of data, since the greater the frequency of certain topics, the greater the knowledge of the topic is likely to be, making it possible to filter out users with different interests and even experts in each field (Barthel et al., 2015).

Several studies attempt to understand communities of attention around different topics or areas of knowledge. For example, Maricato and Manso (2022) studied the identity and characteristics of X users who posted about research from the University of Brasilia. They concluded that individual profiles had a greater impact on research sharing than organizational profiles, demonstrating the need for the latter to increase their visibility and scientific impact on social media. They also found greater engagement in research related to health sciences, medicine, biological sciences, and environmental sciences.

Batooli and Sayyah (2020) evaluated the level of attention that scientific productions on COVID-19 received on social media. They found that the demand for information on the topic in these sources was so high that this research continued to receive attention even after the publication date and concluded that the activities of researchers on social media have the potential to increase the scientific visibility of their work.

Barthel et al. (2015) attempted to identify different groups present on social media and analyze their influence on the correlation with article-level metrics. The groups were categorized according to their fields of interest and experience by analyzing their profile descriptions on X. They concluded that different user groups on social media could influence the scope of metrics in both positive and negative ways. For example, when groups of experts in different fields of knowledge disseminate research, it can lead to a higher rate of views.

Because it is important to understand how articles are discussed and shared by users, Pandian et al. (2019) analyzed different factors that influence the popularity of psychology publications on X. They divided the users who shared the research into academics and non-academics, and experts and non-experts, based on the information available in their profiles. They found that research received more mentions when it was shared by profiles with a larger number of friends, favorites, and lists, by academic users and experts in the field, and by tweets with numerous likes and retweets (reposts of the original publications).

Finally, Ortega (2016) investigated the extent to which a researcher's presence in a social network such as X influences the number of mentions that his or her publications receive. To do so, he sought to answer questions about the users of this network, such as their characteristics and attributes. He concluded that the documents of researchers who use X are shared more than those who do not, and that the number of researchers' followers is the most important metric that influences the retweets of a document.

4 METHODOLOGY

This is an exploratory case study with a quantitative and qualitative approach that aims to understand the network of online care communities that interacted with the most cited scientific article on COVID-19 on X in Brazil between 2020 and 2022. The choice of a single article is based on article level metrics, which also justifies the choice of a qualitative approach, allowing a more in-depth analysis of the profiles.

We chose X, which is commonly used in this type of research, because it is a global channel for information exchange and dissemination of scientific results, with data used by metrics collection systems such as Altmetric.com (Ortega, 2016). This system tracks the attention that research results receive online and is one of the most important altmetric data providers in the world (Joubert; Costas, 2020; Robinson-García et al., 2014).

To select the most shared single scientific article on X in Brazil between 2020 and 2022, the Dimensions database³ was first used to collect publications according to the expressions: "coronavirus" OR "COVID-19" OR "SARS-Cov2". The top ten articles were ranked according to the highest Altmetric scores, and their DOIs were imported into Altmetric.com to see which would be the most shared in Brazil between 2020 and 2022. "Social Distancing Alters the Clinical Course of COVID-19 in Young Adults: A Comparative Cohort Study," published on June 29, 2020, in the journal Clinical Infectious Diseases from the University of Oxford. The study included 4,627 general mentions on public profiles, including original tweets and retweets (republishing a tweet with or without additional commentary).

To identify their community of attention network, Altmetric.com's "mentions" option was used, which provides the number of tweets, mentions data, users, and the direct link to the tweet. The results were filtered between 2020 and 2022 and by the region of Brazil and the source of "original tweets" (tweets and commented retweets). We chose to collect commented retweets because they contain the original content of the reposted tweet and a comment from the user who reposts it. This makes them original tweets that reinforce the connection and discussion of a conversation on social network X (Maleki; Holmberg, 2024).

As a methodological limitation, retweets that did not have a comment in the repost were excluded, leaving a total of 498 posts retrieved after exclusion, of which 15 were discarded because they were deleted or restricted on the social network by the users themselves, leaving 483 publications for analysis.

Once the survey was completed, the data was manually extracted and exported to an Excel spreadsheet with the following information: publication date, user profile (@⁴), name, verified account⁵, bio⁶, emoji⁷ (used in name and bio), location, type of tweet (original or retweet), and link. For the results, the following categories were analyzed: tweets: types and content; and online community of attention network: types of users (verified account, individual, organization); bio (education and occupation); location: city/region of Brazil and emoticons. As in the work of Pulido et al. (2020), the tweets were subjected to an in-depth analysis of their content individually, to be categorized in the research results, considering the meaning intended by the X-user.

Finally, VOSviewer was used to generate word co-occurrence maps of the tweets, as was done in the work of Joubert and Costas (2020) for the co-occurrence of education and occupations and the regions of Brazil. This software helps to visualize data based on clusters (sets of elements - nodes) of network proximity (Van Eck; Waltman, 2022). The size of the nodes represents how many times the keyword occurs, and "co-occurrence seeks to identify elements common to the texts in the database; [...] used to identify significant terms" (Palludeto; Felipini, 2019, p. 318).

5 RESULTS AND DISCUSSION

The most cited article on COVID-19 in X in Brazil was by Bielecki et al. (2020) "Social Distancing Alters the Clinical Course of COVID-19 in Young Adults: A Comparative Cohort Study," which looks at social distancing, the use of masks, and hygiene measures as protocols to reduce the spread of the coronavirus. It presents an outbreak of COVID-19 between March and April 2020 in two similar groups of people in the Swiss Army, before and after social distancing and hygiene measures were implemented. It indicated that the rate of asymptomatic infected soldiers was significantly lower in the group that implemented these measures. In addition to reducing transmission, they were effective in altering the clinical course of the disease in those infected, reducing the rate of symptomatic patients, favoring asymptomatic infection, and inducing a greater immune response.

In Brazil, this international article gained visibility and became the most mentioned COVID-19 research; with 4,627 generals, mentions (tweets and retweets) after biologist Atila Iamarino published it on his personal account⁸. Iamarino became a high-profile figure during the pandemic for providing relevant information about the disease. In his bio on X, he describes himself as "a science communicator and world explainer by choice. According to his Lattes bio⁹, he holds a Ph.D. in biological sciences (microbiology) from the University of São Paulo and three postdoctoral degrees from the Universities of São Paulo and Yale in the United States.

The biologist made five posts mentioning the article, the two most prominent of which were published on July 20, 2020. In the first, he writes: "This result is fantastic and, if replicated, could change the course of COVID-19. They divided the Swiss soldiers into two groups. In the group that followed social distancing, those infected did not develop COVID. In the group without distancing, 30% had complications: D" and adds an image explaining more details about the research¹⁰. In the second: "Masks and distancing can protect you from being infected with the coronavirus and also reduce the chances of having complications from COVID-19. The result is so important that it deserves a video" and offers a three-minute and seven-second video of himself explaining the results of the article¹¹. Both posts received a high number of retweets, likes, and comments from both public and protected profiles and were the main mentions that drove the sharing of this research.

Iamarino's posts, less than a month after the publication of the article by Bielecki et al. (2020), demonstrate the speed offered by this type of sharing, as it allows reaching a wider audience in real time, which consequently increases the number of downloads and citations, as shown by Bik and Goldstein (2013), Bornmann and Haunschild (2018), and Batooli and Sayyah (2020). They also made it possible to highlight the influence of profiles with higher numbers of followers and favorites, as well as of experts in the fields covered, on the increase in mentions, according to Padian et al. (2019). Therefore, Iamarino, who has more than 1,300,000 followers (in 2022) and extensive knowledge in the health field, enabled the research to have a greater reach in Brazil through his posts.

However, although profiles with a larger number of followers have a positive affect the sharing of research (Ortega, 2016), some with a smaller number of followers were also found to interact with the research analyzed, thus increasing the dissemination of the information contained therein. Among those who shared the most, in addition to Iamarino, there were five publications by two other profiles, one with 1,321 followers and the other with 14,495. The first user's posts were all retweets of the original, while the second had three retweets and two original posts. In second place were the profiles that published four tweets, with nine users ranging from 98 to 15,121 followers. Of these, seven only retweeted, while the other two profiles each had three retweets and one original post. These results may be more consistent with the work of Pulido et al. (2020), who found that X users prefer to share retweets that deal with scientifically proven evidence.

5.1 Analysis of tweets: types and content

Of the 4,627 general mentions found on X, 483 were selected for analysis, including original tweets and commented retweets, representing 10.43% of the total. For the results, the following categories were analyzed: tweet types and tweet content.

The 483 publications were divided into original tweets and retweets. The former were considered the original texts that made up the mentions, and the latter were considered reposts. We chose to analyze posts that, in addition to repeating the original tweet, were republished with secondary text, including hashtags [1], emoticons, or graphic characters such as the exclamation point.

There were 34 original tweets (7.03%) and 449 retweets (92.96%). These results are similar to those of Pulido et al. (2020), who showed that users prefer to share scientific evidence through retweets, and Haustein (2018), who also found that most publications related to scientific research are retweets. Thus, among the original tweets, Atila Iamarino's posts stood out the most, followed by Gean Loureiro's tweet [2]. Of the retweets, 166 (34.36%) came from Iamarino's first tweet and 279 (57.76%) from his second, for a total of 445 (92.13%). Loureiro's tweet was republished three times (0.62%)

In a qualitative approach, the content of the tweets was analyzed considering the Brazilian context in the face of COVID-19. It should be noted that the pandemic took place in a period marked by countless controversies about the best treatment and control of the disease in Brazil, when the country faced a vast dissemination of fake news and scientific misinformation by the population, aimed at reducing the perception of the very negative impact of the disease.

Also in this context, according to the work of Pulido et al. (2020), which sought to discover the behavior of the dissemination of disinformation and information based on scientific evidence about COVID-19 in X, the negative effects of false information are of great relevance in the health field, as it can both endanger the health of the population and hinder the implementation of preventive measures by governments. This practice has been facilitated by comments that discredit the credibility of scientific research and can even influence people who discuss the article to learn more about what its results represent. Chart 1 illustrates the public debate on the article analyzed, with some tweets questioning the results, seeking answers to their doubts, highlighting theoretical-methodological aspects, and attacking scientists.

Chart 1
Examples of public debate from the article by Bielecki et al. (2020).

Tweet text	Tweetlink
“We can say then that without distancing; only 30% have had complications. Therefore, if the distancing and quarantine is selective, for risk groups, we lower the lethality even more. Therefore, this shows that quarantine for everyone is not necessary. Right?”	https://twitter.com/jcjuka/status/1285233499750969344
“Number of positive antibody results: 13 out of 154 Average age of participants: 20 (18-28) all healthy and physically active. That is too little and too much bias to say. I guess you don't need that much empirical evidence to support your point, do you?”	https://twitter.com/alexandre_davis/status/1285258352759648256
“If this 'scientist' says it's important, I really have my doubts about the effectiveness of the mask.”	https://twitter.com/Binatti_/status/1285308043920539651
“Do you trust someone without ears? Not me!”	https://twitter.com/fukudaalisson/status/1285345663446847488
“If I understand correctly, I have to join the Swiss army to avoid COVID-19.”	https://twitter.com/rafaelolg/status/1285346906680500226
“I've been saying since the beginning: What if I just take a little mini parsigote from someone with COVID-19 and aspirate it? I'll only have a bit of Corona, and my body will go and kill the virus. Everyone laughed. Nobody encouraged me. The Brazilian "little way" can save us, and nobody pays any attention.”	https://twitter.com/marcelacomunica/status/1286057555035402242
“@oatila Didn't this study say that the amount of virus made a difference in reducing complications? Why is it that if Annita reduces this amount, it doesn't work? Or is it the study about this wrong reduction? I do not doubt it; I'm just confused.”	https://twitter.com/AdoryanB/status/1319819589828681729

• Source: Research data (2022).

However, even with the denial of the events through social networks, the vast majority of tweets analyzed favored the behavior indicated by the World Health Organization (WHO) in confronting the coronavirus through social distancing, wearing a mask, and hygiene measures. Of the 483 tweets, only seven (1.44%) indicated a supposed rejection of the article's findings and questioned its content. Some texts addressed issues that might be contrary to the results of the research, with irony about the results and about Atila Iamarino, while others mentioned measures and drugs that have been shown to be ineffective in the treatment of COVID-19.

To categorize the content of the tweets, the VOSviewer software was used to create networks of co-occurrences of words in the posts. The map was created from the content of 483 tweets. The words or terms with at least three occurrences were considered, resulting in a total of 46 words with the highest prominence and relevance.

Figure 1
Word co-occurrence network of tweets.

The network is made up of seven clusters and 149 links, showing that there was an approximation between the terms of the tweets and similarities in content, with a large proportion citing the words "COVID", which is related to the other clusters, as well as "mask", "complication" and "atila". The terms "science", "wearing a mask", "home", "virus", "importance", and "exposure" were used in relation to the appeal and importance of wearing masks and social distancing ("staying home") to reduce exposure to the virus.

As an example of the use of such terms, some tweets showed Atila Iamarino's attitude towards the pandemic: "Wonderful to see @oatila acting as a science communicator on the one hand and on the other hand to see the group in the comments with cool questions, really cool doubts and in a learning mood. That's influence. May she be an example to us all!"¹² and "If there's one person who has done an award worthy and tireless job, it's @oatila. Total admiration for you."¹³

Others appreciated the science, which has been badly discredited recently: "This is incredible because the research confirms a theory I saw a doctor talk about on TV. Long live science. Follow the thread"¹⁴ and "That's science; wear a mask if you can, keep social distance, and you'll be safer."¹⁵There were also tweets encouraging social distancing, the use of masks, and hygiene measures: "This news is also sensational. Wear a mask, maintain hygiene, and stay home if possible"¹⁶ and "For those who haven't seen it yet, great news about the fight against covid. Masks and distancing may be protecting us more than we thought at the start of the pandemic."¹⁷

Such an analysis of the results reinforces the conclusions of Pulido et al. (2020) that a deeper understanding of how the care community network has behaved on a given issue helps authorities to understand what information is being shared. In this way, they can seek to combat misinformation and revalidate scientifically proven data, with social media, such as institutional profiles, being important tools for sharing and verifying information.

5.2 Analysis of the online care community network: types of users, bio, location, and emojis

For the analysis of the community of attention network in the article by Bielecki et al. (2020), we considered the categories of user type (verified account, individual, and organization); bio description (education and occupation); location: city/state/region of Brazil and emojis (used in name and bio). By studying the 483 tweets and retweets, it was possible to identify the profiles of X users who published them, 13 of which were repeated profiles, considering 470 different users.

For the category of user types, we first looked at the verified profiles that mentioned the article. As public figures, these users have a great influence on their audience. We found 12 accounts (2.55%), shown in Chart 2, along with the number of followers, as was also done in Ortega's research (2016).

Chart 2
Verified accounts

User name	@	Bio	No. of followers
Atila Iamarino	@oatila	Science communicator and world explainer by choice [e-mail] On Hiperconectado on TV Cultura: Saturdays at 8:30 p.m. and Wednesdays at 8 p.m.	1.308.091
Mídia NINJA	@MidiaNINJA	Independent Narratives, Journalism, and Action.	1.112.334
Biodiversidade Brasileira	@BiodiversidadeB	Welcome to Brazil's largest biodiversity profile | By João Pedro Salgado | [e-mail]	339.071
fabiula nascimento	@fabiunascimento	Actress Fabiula Nascimento's official profile updated by herself	144.762
Ethel Maciel, PhD	@EthelMaciel	Full Professor @ufesoficial Nurse. PhD in Epidemiology. Researcher. Elected Rector of @ufesoficial and not appointed. President @redetuberculosis	71.918
Gean Loureiro	@GeanLoureiro	Candidate for Governor of SC. My number is 44. Married and the father of 4 girls, twice mayor of Florianópolis and very proud to be from Santa Catarina.	47.694
ganga	@thabataganga	Technology Scientist, Biomedical Engineer and Master's student in Technological Innovation | contact: [e-mail]	27.752
Nunca vi 1 cientista	@_NV1C	Here you'll find credible, humorous information and a safe place to ask questions :) email: [e-mail]	26.250
Chico D'Angelo	@chico_dangelo	Physician, Federal Deputy (PDT-RJ), former Secretary of Health of Niterói and former President of the Culture Commission of the Chamber of Deputies.	17.944
Bruno Natal (RESUMIDO)	@URBe	Journalist and host of RESUMIDO, talking about the impact of technology on our lives | #1 on Apple Podcasts | http://resumido.cc | http://youtube.com/resumido	13.736
Mel Markoski	@melmarkoski	Biologist, master's and doctorate in Cellular and Molecular Biology, post-doctorate in Immunology and Cancer. Professor of Biosafety at UFCSPA. Member of the COVID-19 Analysis Network	7.792
Maria Leticia	@MedicaLeticia	Coroner, Gynecologist, Councilwoman in Curitiba (PV), CMC Women's Attorney, President of the Environment Commission	1.792

• Source: Research data (2022).
• Note: the number of followers was verified on December 28, 2022.

There were several verified accounts, including science communicators, journalism profiles and journalists, doctors, politicians, scientists, teachers, and artists. The number of followers shows that most of these profiles have a high level of engagement with the general public, which may be an indication of the wide reach of their content, similar to the results of Padian et al. (2019) and Ortega (2016). For the latter author, the network of followers of those who share scientific research allows them to multiply their audience and increase their number of retweets.

In the second stage, user profiles were divided into organizational and individual, as done by Maricato and Manso (2022). For the former, profiles representing institutions, organizations, groups of people with common interests or trademark profiles, 10 (2.12%) were found, while for the latter, individual accounts (personal profiles), 460 (97.87%) were found. These results are similar to those of Maricato and Manso (2022), in which the vast majority were individual profiles.

The organizational ones were profiles of a Brazilian scientific journal, a public university laboratory, independent journalism, profiles on Brazilian biodiversity and environmental monitoring (both run by independent people with knowledge of the areas), an agency and a production company, as well as representatives of a personal blog, a website on technology, the dissemination of science and the lives of scientists and a podcast about world events. Some of these results corroborate those of Maricato and Manso (2022), who also noted the relevance of the engagement of profiles from universities, departments, and journals in the dissemination of science.

The 460 individual profiles were analyzed based on the information extracted from each one's bio, in which characteristics such as education and occupation were noted, as was done by Barthel et al. (2015). 282 (61.30%) did not provide such data, and it was possible to analyze 178 (38.69%) profiles that largely presented more than one type of education and occupation, as also noted by Haustein (2018). In addition to those already mentioned in the verified accounts, there was a predominance of profiles related to academia and science, similar to the results of Haustein (2018) and Maricato and Manso (2022). Therefore, profiles were found for: students (40), professors (18), masters/master's students (15), doctors/doctoral students (14), scientists (eight), researchers (five), and educators (three).

Figure 2
Co-occurrence network of schooling and occupation.

In general, the professions of journalist (18), designer (11), lawyer (eight), doctor (seven), illustrator (seven), physiotherapist (six), administrator (six), photographer (six), psychologist (five) and engineer (five) also stand out. Activists from various causes, such as the environment and social minorities, presented six profiles according to the descriptions, and profiles of Brazilian politicians totaled five.

The predominance of the individual profile user type, according to Maricato and Manso's (2022) results, tends to have a greater impact on the culture of promoting science outside the academic environment than the organizational profiles, in which one can consider greater individual motivation than an organizational initiative in the dissemination of scientific results. Allied to this is the interest of people from different areas, not just academia, in sharing issues related to the health of the population, as also demonstrated in the research by Pandian et al. (2019) and Maricato and Manso (2022), in which fields such as medicine and health receive greater attention online, indicating great interest in the general public.

For the category of location in Brazil from which users originate, it is important to clarify that the X user can freely choose what to write in this option, be it no information, their city, their state, their country, or even invented data, as pointed out by Haustein (2018). We therefore opted to use credible information about the city, state, and country.

From the 470 organizational and individual profiles, 446 pieces of information on locations were found, 48 of which are repeated data in which more than one user originates from that region. A density map was created in VOSviewer of the co-occurrence of the information obtained between cities, states, and countries.

Figura 3
Density network of site co-occurrence in Brazil.

It is noteworthy that 59 users only indicated “Brazil”, not their state or city. Among the states, 15 were different, such as São Paulo and Rio de Janeiro, which could be either the state or the capital. And 89 cities, such as the capitals São Paulo and Rio de Janeiro, as well as Porto Alegre, Brasília, Fortaleza, Belo Horizonte and Florianópolis, which also occurred in greater quantities. The general distribution by region of Brazil:

Table 1
Distribution of user profiles by region of Brazil.

Region/country	No. of cities	No. of states	No. of profiles	% of profiles
North	7	2	24	5,38
Northeast	15	6	66	14,79
Midwest	5	1	24	5,38
Southeast	38	3	201	45,06
South	24	3	72	16,14
Brazil	-	-	59	13,22
Total	89	15	446	100

• Source: Research data (2022).

There is a strong predominance of profiles from the southeast of the country. Based on the content of the tweets, it is not possible to say why the article was shared so differently between the regions of Brazil. However, July 2020, the date on which the highest number of shares occurred, was the deadliest month in the country so far, and the state of São Paulo was the hardest hit (Welle, 2020), which could have influenced the greater promotion of the research in this state and, consequently, in the southeast region.

For the last category analyzed, emoticons are increasingly studied by academics and are considered as independent and nonverbal forms of online communication (Bai et al., 2019). Their understanding and use can be influenced by the individual characteristics of users, their cultural backgrounds, and their linguistic and social contexts, leading to different interpretations of their meanings. They can express emotions, moods, people, animals, food, activities, actions, movements, and body parts, as well as international objects and flags (Bai et al., 2019; Das; Wiener; Kareklas, 2019; Eisner et al., 2016). Based on these characteristics, the most commonly used emoticons in names and biographies were investigated.

First, the Brazilian flag emoji was used in 36 profiles. This image has different meanings in the country from a semiotic perspective. Among other things, it may have been used to express the country of origin, support for the Brazilian men's soccer team in the 2022 FIFA World Cup (a period that coincides with the data collection of this research), and also as an identity issue that has gained great appeal among the Brazilian population supporting right-wing parties and politicians recently. However, there has been an attempt to revive its use by the part of the population that does not identify with this political wing, which may justify the high use of this emoji in the profiles analyzed.

The second most used emoji was the one representing a red triangular flag, which appeared in 16 profiles. One of its uses can be exemplified by the support for the president-elect in 2022, Luiz Inácio Lula da Silva, as it was also used in some profiles, along with other emojis that can represent characteristics of this support, such as the numbers one and three, which form 13, the number of the president's party, as well as the animal Lula, which also represents him in the Brazilian context. Some of these profiles also used the Brazilian flag as a form of support for President Lula.

Finally, 26 profiles used emoticons representing science and academia. These images ranged from depictions of students, researchers, professors, and graduates wearing capes, to images of open and stacked books, computers, telescopes and microscopes, graphs and tables, and images of DNA, laboratory instruments, medicines, and syringes, the latter of which could also represent COVID-19 vaccines.

6 CONCLUSION

The COVID-19 pandemic has changed society worldwide. Protocols such as social distancing, wearing masks, and personal hygiene measures have become part of people's daily lives. As the severity of the disease, increased, scientific productions on these topics were published and shared on social media to reach a wider audience.

Altmetrics and article-level metrics make it possible to get to know the online communities of attention that interact with scientific works published on these types of web sources, including X. In this context, the aim of this study was to get to know the community of attention that interacted with the most mentioned scientific article on COVID-19 on X in Brazil, entitled "Social Distancing Alters the Clinical Course of COVID-19 in Young Adults: A Comparative Cohort Study" by Bielecki et al. (2020).

As a result, 4,627 general mentions of X were found, of which 483 were selected for analysis, including original tweets (7.03%) and commented retweets (92.96%). Among the originals, there was a lot of interaction with two posts by biologist Atila Iamarino, which together received 92.13% of the retweets. As for the content of the tweets, there were seven possible disagreements with the results of the article, while the vast majority were in favor. Topics such as wearing a mask, social distancing, exposure to the virus, and science were the most frequently mentioned in the texts.

To understand the community of attention, 470 different X users were found, with 12 verified accounts. They were also divided into individual (97.87%) and organizational (2.12%) profiles. These results indicate that X users are concerned with sharing legitimate, science-based information about health and protection against COVID-19, with the aim of reaching a wider audience and providing and promoting knowledge through their posts.

Among the individuals, profiles related to academia and science stood out. Among the regions of Brazil with the most users, the Southeast, with Rio de Janeiro and São Paulo standing out. Finally, the most used emoji included the Brazilian flag, a red triangular flag, and drawings representing science and academia, including the syringe emoji, which may represent support for the COVID-19 vaccine.

With only the source of X as a limitation, future work on the same theme in other web sources is suggested, as well as on different themes. It is necessary to be increasingly aware of online communities of attention that interact with scientific works from different fields of knowledge that are disseminated virtually.

Acknowledgments:

Not applicable.

References

• ADIE, E.; ROE, W. Altmetric: enriching scholarly content with article‐level discussion and metrics. Learned Publishing, London, v. 26, n. 1, p. 11-17, jan. 2013. DOI: https://doi.org/10.1087/20130103
  » https://doi.org/10.1087/20130103
• AMARAL, H. O papel das tvs públicas no contexto da pandemia de covid-19: análise noticiosa da JFTV Câmara. In: CORDEIRO, D. F. et al. (org.). Mídias, informação e ciência de dados: pesquisas, tendências e interfaces. Goiânia: Cegraf UFG, 2020. cap. 16, p. 237-251.
• ARAÚJO, R. F de. Communities of attention networks: introducing qualitative and conversational perspectives for altmetrics. Scientometrics, Hungary, v. 124, p. 1793-1809, jun. 2020. DOI: https://doi.org/10.1007/s11192-020-03566-7
  » https://doi.org/10.1007/s11192-020-03566-7
• ARAÚJO, R. F. de et al. O debate público e a desinformação científica: estudo sobre a nitazoxanida como tratamento para COVID-19. In: ENCONTRO NACIONAL DE PESQUISA EM CIÊNCIA DA INFORMAÇÃO, 23., 2023, Aracaju. Anais [...]. Aracaju: Universidade Estadual de Sergipe, 2023. p. 1-11.
• BAI, Q. et al. A systematic review of emoji: current research and future perspectives. Frontiers in Psychology, Switzerland, v.10, n. 2.221, p. 1-16, oct. 2019. DOI: https://doi.org/10.3389/fpsyg.2019.02221
  » https://doi.org/10.3389/fpsyg.2019.02221
• BARTHEL, S. et al. What does twitter measure? influence of diverse user groups in altmetrics. In: PROCEEDINGS OF THE 15TH ACM/IEEE-CS JOINT CONFERENCE ON DIGITAL LIBRARIES, 15., 2015. Anais [...]. New York, USA: Association for Computing Machinery, 2015. p. 119-128.
• BATOOLI, Z.; SAYYAH, M. Measuring social media attention of scientific research on Novel Coronavirus Disease 2019 (COVID-19): an investigation on article-level metrics data of Dimensions, Preprint, [S.l.], 14 apr. 2020. DOI: https://doi.org/10.21203/rs.3.rs-21980/v1
  » https://doi.org/10.21203/rs.3.rs-21980/v1
• BIELECKI, M. et al. Social distancing alters the clinical course of COVID-19 in young adults: a comparative cohort study. Clinical Infectious Diseases, Belgium, v. 72, n. 4, p. 598-603, jun. 2020. DOI: https://doi.org/10.1093/cid/ciaa889
  » https://doi.org/10.1093/cid/ciaa889
• BIK, H. M.; GOLDSTEIN, M. C. An introduction to social media for scientists. PLoS Biology, United States, v. 11, n. 4, p. 1-8, apr. 2013. DOI: https://doi.org/10.1371/journal.pbio.1001535
  » https://doi.org/10.1371/journal.pbio.1001535
• BORNMANN, L.; HAUNSCHILD, R. Alternative article-level metrics: the use of alternative metrics in research evaluation. EMBO Reports, London, v. 19, p. 1-4, nov. 2018. DOI: https://doi.org/10.15252/embr.201847260
  » https://doi.org/10.15252/embr.201847260
• DAS, A. K.; MISHRA, S. Genesis of altmetrics or article-level metrics for measuring efficacy of scholarly communications: Current perspectives? Journal of Scientometric Research, India, v. 3, n. 2, p. 82-92, 2014. DOI: https://doi.org/10.48550/arXiv.1408.0090
  » https://doi.org/10.48550/arXiv.1408.0090
• DAS, G.; WIENER, H. J. D.; KAREKLAS, I. To emoji or not to emoji? examining the influence of emoji on consumer reactions to advertising. Journal of Business Research, Papua New Guinea, v. 96, p. 147-156, 2019. DOI: https://doi.org/10.1016/j.jbusres.2018.11.007
  » https://doi.org/10.1016/j.jbusres.2018.11.007
• EISNER, B. et al. emoji2vec: Learning emoji representations from their description. In: PROCEEDINGS OF THE 4TH INTERNATIONAL WORKSHOP ON NATURAL LANGUAGE PROCESSING FOR SOCIAL MEDIA AT EMNLP 2016, 4., 2016. Anais [...]. Austin, USA: Association for Computational Linguistics, 2016. p. 1-7.
• GOUVEIA, F. C. Altmetria: métricas de produção científica para além das citações. Liinc Em Revista, Rio de Janeiro, v. 9, n. 1, p. 214-227, 2013. DOI: https://doi.org/10.18617/liinc.v9i1.569
  » https://doi.org/10.18617/liinc.v9i1.569
• HAUSTEIN, S. Scholarly Twitter metrics: how, when and what does the Twittersphere tweet about science? In: BRAZILIAN MEETING OF BIBLIOMETRICS AND SCIENTOMETRICS, 6., 2018. Anais [...]. Rio de Janeiro: Universidade Federal do Rio de Janeiro, 2018.
• HAUSTEIN, S. et al. Tweeting biomedicine: an analysis of tweets and citations in the biomedical literature. Journal of the Association for Information Science and Technology, United States, v. 65, n. 4, p. 656-669, 2014. DOI: https://doi.org/10.1002/asi.23101
  » https://doi.org/10.1002/asi.23101
• HISING, E. Rede social X é suspensa no Brasil após ordem de Moraes. G1, [S.l.], 31 ago. 2024. Disponível em: https://l1nk.dev/XggVl Acesso em: 26 jul. 2022.
  » https://l1nk.dev/XggVl
• HOLMBERG, K. et al. Astrophysicists’ conversational connections on Twitter. PLoS ONE, San Francisco, v. 9, n. 8, p. 1-13, aug. 2014. DOI: https://doi.org/doi:10.1371/journal.pone.0106086
  » https://doi.org/doi:10.1371/journal.pone.0106086
• JOUBERT, M.; COSTAS, R. Getting to know science tweeters: a pilot analysis of south African Twitter users tweeting about research articles. Journal of Altmetrics, United States, v. 2, n. 1, p. 1-14, 2020. DOI: https://doi.org/10.29024/joa.8
  » https://doi.org/10.29024/joa.8
• LEAL, M. R. C.; RIBEIRO, G. M. de C. Disputas discursivas em torno da cloroquina / hidroxicloroquina no combate ao coronavírus no Brasil: entre a ciência e a desinformação. In: CORDEIRO, D. F. et al. (org.). Mídias, informação e ciência de dados: pesquisas, tendências e interfaces. Goiânia : Cegraf UFG, 2020. cap. 13, p. 188-207.
• LIU, J.; ADIE, E. New perspectives on article-level metrics: developing ways to assess research uptake and impact online. Insights, United Kingdom, v. 27, n. 2, p. 153-158, 2013. DOI: http://dx.doi.org/10.1629/2048-7754
  » http://dx.doi.org/10.1629/2048-7754
• MALEKI, A.; HOLMBERG, K. Tweeting and retweeting scientific articles: implications for altmetrics. Scientometrics, Hungary, v.129, p. 6197-6220, 2024. DOI: https://doi.org/10.1007/s11192-024-05127-8
  » https://doi.org/10.1007/s11192-024-05127-8
• MARICATO, J. M.; MANSO, B. L. C. Characterization of the communities of attention interacting with scientifc papers on Twitter: altmetric analysis of a Brazilian University. Scientometrics, Hungary, v. 127, p. 3815-3835, jun. 2022. DOI: https://doi.org/10.1007/s11192-022-04442-2
  » https://doi.org/10.1007/s11192-022-04442-2
• MENDES, M. M.; MARICATO, J. de M. Das apresentações públicas às redes sociais: : apontamentos sobre divulgação científica na mídia brasileira. Comunicação & Informação, Goiânia, Goiás, v. 23, p. 1-16, 2020. DOI: https://doi.org/10.5216/ci.v23i.49959
  » https://doi.org/10.5216/ci.v23i.49959
• ORDUÑA-MALEA, E.; MARTÍN-MARTÍN, A.; DELGADO-LÓPEZ-CÓZAR, E. The next bibliometrics: almetrics (author level metrics) and the multiple faces of author impact. El profesional de la información, Spain, v. 25, n. 3, p. 485-470, may/jun. 2016. DOI: https://doi.org/10.3145/epi.2016.may.18
  » https://doi.org/10.3145/epi.2016.may.18
• ORTEGA, J. L. To be or not to be on Twitter, and its relationship with the tweeting and citation of research papers. Scientometrics, Hungary, v. 109, p. 1353-1364, aug. 2016. DOI: https://doi.org/10.1007/s11192-016-2113-0
  » https://doi.org/10.1007/s11192-016-2113-0
• PALLUDETO, A. W. A.; FELIPINI, A. R. Panorama da literatura sobre a financeirização (1992-2017): uma abordagem bibliométrica. Econ. e Soc., Campinas, v. 28, n. 2, p. 313-337, maio/ago. 2019. Disponível em: https://periodicos.sbu.unicamp.br/ojs/index.php/ecos/article/ view/8656961 Acesso em: 03 jul. 2022.
  » https://periodicos.sbu.unicamp.br/ojs/index.php/ecos/article/ view/8656961
• PANDIAN, N. D. S. et al. Altmetrics: factor analysis for assessing the popularity of research articles on Twitter. Journal Of Information Science Theory And Practice, Republic of Korea, v. 7, n. 4, p. 33-44, 2019. DOI: https://doi.org/10.1633/JISTaP.2019.7.4.3
  » https://doi.org/10.1633/JISTaP.2019.7.4.3
• PEIXOTO, P. N.; ARAÚJO, R. F. de. Práticas informacionais na web social e os influenciadores digitais: uma análise espaço-temporal dos YouTubers brasileiros. Folha de Rosto, Juazeiro do Norte, v. 8, n. 1, p. 162-187, abr. 2022. DOI: https://doi.org/10.46902/2022n1p162-187
  » https://doi.org/10.46902/2022n1p162-187
• PULIDO, C. M. et al. COVID-19 infodemic: more retweets for science-based information on coronavirus than for false information. International Sociology, United Kingdom, v. 35, n. 4, p. 377-392, 2020. DOI: https://doi.org/10.1177/026858092091475
  » https://doi.org/10.1177/026858092091475
• ROBINSON-GARCÍA, N. et al. New data, new possibilities: exploring the insides of Altmetric.com. El profesional de la información, Spain, v. 23, n. 4, p. 359-366. 2014. DOI: https://doi.org/10.48550/arXiv.1408.0135
  » https://doi.org/10.48550/arXiv.1408.0135
• SHAMSI, A.; LUND, B. D.; SEYYEDHOSSEINI, S. Sharing of retracted COVID-19 articles: an altmetric study. Journal of the Medical Library Association, United States, v. 110, n. 1, p. 97-102, jan. 2022. DOI: https://doi.org/10.5195/jmla.2022.1269
  » https://doi.org/10.5195/jmla.2022.1269
• SILVA, M. M.; CENDON, B. V. Análise da produção e divulgação de desinformação científica no meio digital. Pesquisa Brasileira em Ciência da Informação e Biblioteconomia, João Pessoa, v. 17, n. 4, 2022. Disponível em: https://brapci.inf.br/#/v/245453 Acesso em: 20 ago. 2024.
  » https://brapci.inf.br/#/v/245453
• SUGIMOTO, C. R. et al. Scholarly use of social media and altmetrics: a review of the literature. Journal of the Association for Information Science and Technology, United States, v. 68, p. 2037-2062, 2017. DOI: https://doi.org/10.1002/asi.23833
  » https://doi.org/10.1002/asi.23833
• THELWALL, M.; WILSON, P. Mendeley readership altmetrics for medical articles: An analysis of 45 fields. Journal of the Association for Information Science and Technology, United States, v. 67, p. 1962-1972. 2015. DOI: https://doi.org/10.1002/asi.23501
  » https://doi.org/10.1002/asi.23501
• TRAJKOVSKI, V. The role of article level metrics in scientific publishing. The Journal of Special Education and Rehabilitation, Skopje, Republic of Macedonia, v.17, n. 1-2, p. 5-15, 2016. DOI: https://doi.org/10.19057/jser.2016.0
  » https://doi.org/10.19057/jser.2016.0
• X Corp. Glossario. 2024. Disponível em: https://help.x.com/pt/resources/glossary Acesso em: 27 nov. 2024.
  » https://help.x.com/pt/resources/glossary
• VAN ECK, N. J.; WALTMAN, L. VOSviewer Manual. Universiteit Leiden, 2022. p. 1-52.
• VANTI, N. A. P.; SANZ-CASADO, E. Altmetria: a métrica social a serviço de uma ciência mais democrática. Transinformação, Campinas, v. 28, n. 3, p. 349-358, dez. 2016. DOI: https://doi.org/10.1590/2318-08892016000300009
  » https://doi.org/10.1590/2318-08892016000300009
• WELLE, D. Julho foi mês mais fatal da pandemia de covid-19 no Brasil: foram quase 33.000 mortes SP foi o Estado mais atingido... Poder 360. [S.l.], 2 ago. 2020. Disponível em: https://encr.pw/RcDLs Acesso em: 3 jul. 2022.
  » https://encr.pw/RcDLs
• YOUSEFIFINAGHANIA, S. et al. An analysis of COVID-19 vaccine sentiments and opinions on Twitter. International Journal of Infectious Diseases, Netherlands, v. 108, p. 256-262, 2022. DOI: https://doi.org/10.1016/j.ijid.2021.05.059
  » https://doi.org/10.1016/j.ijid.2021.05.059

• 

  
  

• 

  
  

• JITA:
  BB. Bibliometric methods
• ODS:
  3. Health and Wellness
• Funding:
  Not applicable.
• Ethical approval:
  Not applicable.
• Availability of data and material:
  Not applicable.
• Image:
  Photo extracted from the Lattes Curriculum.

• 1
  In 2023, the social network “Twitter” was renamed “X” after being bought by entrepreneur Elon Musk.
• 2
  Available at: https://www.altmetric.com/.
• 3
  Available at: https://www.dimensions.ai/.
• 4
  “The @ sign is used to call out usernames in posts: ”Hi @X!” People use your username to mention you in posts, send you a message or link to your profile” (X Corp., 2024).
• 5
  “A verified X account receives a blue verification icon to indicate that the creator of those posts is a legitimate source. Verified accounts include public figures and possible victims of impersonation on X” (X Corp., 2024).
• 6
  “Your bio is a short personal description (up to 160 characters) that appears on your profile and serves to characterize your identity on X” (X Corp., 2024).
• 7
  “An X emoji is a specific series of letters immediately preceded by the # sign that generates an icon on X. For example, a national flag or other small image (X Corp., 2024).
• 8
  Available at: https://twitter.com/oatila.
• 9
  Available at: http://lattes.cnpq.br/4978322672579487.
• 10
  Available at: https://twitter.com/oatila/status/1285228826398789632.
• 11
  Available at: https://twitter.com/oatila/status/1285255942800637952.
• 12
  Available at: https://twitter.com/wesleybrasil/status/1285231280360239104.
• 13
  Available at: https://twitter.com/Renataftavares/status/1285282344660873218.
• 14
  Available at: https://twitter.com/MarcianoBrito13/status/1285232119271370756.
• 15
  Available at: https://twitter.com/marceloprr1/status/1285264937875574790.
• 16
  Available at: https://twitter.com/robertafalcao/status/1285236282562674688.
• 17
  Available at: https://twitter.com/fidefrancisco/status/1286424693139034112.

• Editor:
  Gildenir Carolino Santos

Data availability

Not applicable.

Publication Dates

• Publication in this collection
  03 Feb 2025
• Date of issue
  2025

History

• Received
  17 May 2024
• Accepted
  05 Nov 2024
• Published
  04 Dec 2024