Coronavirus Goes Viral: Quantifying the COVID-19 Misinformation Epidemic on Twitter.

Kouzy R, Abi Jaoude J, Kraitem A, El Alam MB, Karam B, Adib E, Zarka J, Traboulsi C, Akl EW, Baddour K. · 2020

Read Coronavirus Goes Viral: Quantifying the COVID-19 Misinformation Epidemic on Twitter. by Kouzy R, Abi Jaoude J, Kraitem A, El Alam MB, Karam B, Adib E, Zarka J, Traboulsi C, Akl EW, Baddour K. (2020) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc7152572.txt.

Introduction

Kouzy and co-authors measure COVID-19 misinformation on Twitter, in Cureus. They searched 14 trending hashtags and keywords about the epidemic and checked 673 tweets against verified and peer-reviewed sources. Most tweets came from informal individuals or groups (66%), and 19.2% came from verified accounts.

Overall, 81.4% of tweets carried genuine information, but 24.8% included misinformation and 17.4% included unverifiable claims. Informal accounts posted misinformation at 33.8%, and unverified accounts at 31.0% against 12.6% for verified ones. Health care and public health accounts had the least unverifiable content. Likes and retweets did not distinguish true from false posts, a warning about judging by popularity.

Catalog record

CREATORKouzy R, Abi Jaoude J, Kraitem A, El Alam MB, Karam B, Adib E, Zarka J, Traboulsi C, Akl EW, Baddour K.
YEAR2020
COLLECTIONPAPERS
FORMATTXT
FILE SIZE17 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC7152572

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Coronavirus Goes Viral: Quantifying the COVID-19 Misinformation Epidemic on Twitter By Ramez Kouzy, Joseph Abi Jaoude, Afif Kraitem, Molly B El Alam, Basil Karam, Elio Adib, Jabra Zarka, Cindy Traboulsi, Elie W Akl, Khalil Baddour, Alexander Muacevic, John R Adler Published in: Cureus DOI: 10.7759/cureus.7255 ABSTRACT Abstract Background Since the beginning of the coronavirus disease 2019 (COVID-19) epidemic, misinformation has been spreading uninhibited over traditional and social media at a rapid pace. We sought to analyze the magnitude of misinformation that is being spread on Twitter (Twitter, Inc., San Francisco, CA) regarding the coronavirus epidemic. Materials and methods We conducted a search on Twitter using 14 different trending hashtags and keywords related to the COVID-19 epidemic. We then summarized and assessed individual tweets for misinformation in comparison to verified and peer-reviewed resources. Descriptive statistics were used to compare terms and hashtags, and to identify individual tweets and account characteristics. Results The study included 673 tweets. Most tweets were posted by informal individuals/groups (66%), and 129 (19.2%) belonged to verified Twitter accounts. The majority of included tweets contained serious content (91.2%); 548 tweets (81.4%) included genuine information pertaining to the COVID-19 epidemic. Around 70% of the tweets tackled medical/public health information, while the others were pertaining to sociopolitical and financial factors. In total, 153 tweets (24.8%) included misinformation, and 107 (17.4%) included unverifiable information regarding the COVID-19 epidemic. The rate of misinformation was higher among informal individual/group accounts (33.8%, p: <0.001). Tweets from unverified Twitter accounts contained more misinformation (31.0% vs 12.6% for verified accounts, p: <0.001). Tweets from healthcare/public health accounts had the lowest rate of unverifiable information (12.3%, p: 0.04). The number of likes and retweets per tweet was not associated with a difference in either false or unverifiable content. The keyword “COVID-19” had the lowest rate of misinformation and unverifiable information, while the keywords “#2019_ncov” and “Corona” were associated with the highest amount of misinformation and unverifiable content respectively. Conclusions Medical misinformation and unverifiable content pertaining to the global COVID-19 epidemic are being propagated at an alarming rate on social media. We provide an early quantification of the magnitude of misinformation spread and highlight the importance of early interventions in order to curb this phenomenon that endangers public safety at a time when awareness and appropriate preventive actions are paramount. Keywords: coronavirus, twitter, social media, epidemic, public health, pandemic, infodemic, covid-19 INTRODUCTION Since December 2019, the coronavirus disease 2019 (COVID-19) epidemic has swept the world, causing significant burden and an increasing number of hospitalizations [1,2]. While public health and healthcare officials rushed to identify and contain the spread of the virus, information was spreading uninhibited over traditional and social media platforms at a strikingly rapid pace. Both the impact of the disease and the lack of information associated with it allowed medical misinformation to rapidly surface and propagate on various social media platforms. Previous reports have highlighted a similar trend during recent public health emergencies, mainly the Ebola and Zika outbreaks [3,4]. Such a phenomenon is alarming on both individual and public health levels to an extent that governing bodies are realizing its gravity and attempting to limit its effects [5-7]. Misinformation can be defined as a “claim of fact that is currently false due to lack of scientific evidence” [5]. It propagates without constraints, does not entail any curation or peer-review, and does not require any professional verifications. This makes it ideal to spread on social media and become amplified by the information silos and echo chambers of personally tailored content, particularly during times of public tension like the current COVID-19 epidemic [8]. To our knowledge, attempts to quantify misinformation during the current COVID-19 epidemic are still lacking. Hence, in this report, we seek to analyze the magnitude of misinformation that is being spread on Twitter (Twitter, Inc., San Francisco, CA) regarding the coronavirus epidemic. MATERIALS AND METHODS Data collection We performed an online search of the Twitter social media platform on February 27, 2020. We used the Twitter Archiver add-on to search Twitter for tweets containing one or more of 11 common hashtags and three common key terms pertaining to the COVID-19 epidemic that were identified by the Symplur (Symplur LLC, Los Angeles, CA) analytical tool (Figure 1). Our search was limited to tweets in the English language and to those that initially received at least five retweets. We excluded tweets that had four or fewer retweets. We selected a random sample of 50 tweets from search terms, which yielded more than 100 tweets that fit our inclusion criteria. Samples were selected based on computer-generated random sequences. [Figure: Details of the most common hashtags and search terms pertaining to the COVID-19 epidemic] For every individual tweet, a set of predetermined variables were collected. The different tweet/account characteristics were either automatically extracted by Twitter Archiver or manually collected from the tweets on Twitter by the authors. All data were publicly available, and hence this study did not require institutional review board approval. User accounts were classified based on content into the following categories: informal individual/group, business/NGO/government, news outlet/journalist, and healthcare/public health/medical (Table 1). Accounts’ verified status was additionally noted. A verified account is defined as one of public interest that is deemed to be authentic by Twitter. [Table: Twitter account characteristics] Tweets were categorized based on content tone into the following categories: serious, humorous, and opinions. Tweets labeled as serious were those with information pertaining to COVID-19 or revolving around it, while humorous tweets consisted of jokes or memes. Tweets labeled as opinions were posts that conveyed the account's viewpoint and did not relay any novel information. Tweets were further classified based on content type into medical/public health, financial, and/or sociopolitical. Tweets that contained genuine information regarding the COVID-19 epidemic were identified. Such information was cross-matched with the information presented by the World Health Organization (WHO), the Center for Disease Control and Prevention (CDC), peer-reviewed scientific journals, and prominent news outlets [9-12]. Tweets that included information that could be clearly refuted using one of the above-mentioned references were considered under misinformation. Tweets that could not be proven correct or incorrect by the references were designated as unverifiable information. Statistical analysis Descriptive statistics were conducted to analyze the Twitter accounts and tweets’ characteristics. Bar graphs were generated using Microsoft Office Excel version 16 (Microsoft Corporation, Redmond, WA). Chi-square statistic was used to calculate p-values for the association between account/tweet characteristics and the presence of misinformation or unverifiable information. Statistical significance was set a priori at a two-sided p-value of 0.05. All analyses were performed using IBM SPSS Statistics Version 26 (IBM Corporation, Armonk, NY). RESULTS Account and tweet characteristics A total of 673 tweets were included and analyzed in this study. Most tweets were posted by informal individuals or groups (448, 66.6%), followed by news outlets or journalists (111, 16.5%). Of all accounts, 129 (19.2%) were Twitter verified accounts. Table 2 presents the characteristics of the tweets analyzed. The majority of tweets included serious content (614, 91.2%), with information pertaining to the COVID-19 epidemic (548, 81.4%), and only 41 tweets (6.1%) included humorous content. The most frequent topic was medical/public health (468, 69.5%), followed by sociopolitical (242, 40.0%) and financial (38, 5.6%). [Table: Individual tweet characteristics] Misinformation and

Read the complete text