The Effects of the COVID-19 “Infodemic” on Journalistic Content and News Feed in Online and Offline Communication Spaces

Abstract

The systematic coverage of the coronavirus pandemic by the Greek mass media began in February 2020, specifically, from the time the virus made its appearance in the most significant way in Italy. Until then, news about the virus had been sporadically visible depending mainly on news reports coming from the international media and press agencies. The assessment of the COVID-19 pandemic as an “infodemic” by the World Health Organization (WHO) made obvious the need to study media coverage and map its patterns, along with the unprecedented political and social response and the massive consequences on the global economy. Through a large content analysis, containing 7457 news items from 13 different media outlets, plus a comparative Twitter analysis of 36,317 tweets, we took the present situation as an opportunity to collect real-time data but also as a point of departure for addressing issues connected to journalistic practices and technological changes in the framework of COVID-19. According to our findings, the Greek media faced the crisis “with a view to the world”, emphasizing international coverage, giving priority to the authorities and scientists, and keeping (at least in their majority) hoaxes and conspiracy theories out of the agenda.

1. Introduction

Digital technologies and social media platforms have produced transformational changes in journalism. Several aspects of newsroom work have already been sufficiently documented (Vu et al. 2020; Lecheler and Kruikemeier 2016; Hellmueller et al. 2013), providing further insight into how journalism is practiced today. Much of the scholarly work focuses on the interaction between established norms and journalistic standards, such as impartiality, and the new imperatives of social media.

At the onset of the COVID-19 pandemic, at the beginning of 2020, it became clear that due to the nature of the pandemic and the measures taken, online spaces of information would gain an even bigger advantage compared to offline ones. At the same time, the role of authoritative sources in providing the dominant frames has been questioned, along with the capacity of media personnel to provide verified information.

Therefore, the pandemic provides us a unique occasion to study media coverage in both online and offline spaces, in the framework of the often-overlooked relationship between journalism and public health (Bernadas and Ilagan 2020), and at the same time monitor the related debate on Twitter, a medium with different technological affordances that “privileges the rapid production and dissemination of fragments of information in a public, networked environment, resulting in an expectation of speedy and pithy textual exchanges” (Hermida and Mellado 2020, p. 872). Greek print and digital media, along with the Greek Twitter, were used as a case study in an attempt to include the factor of national particularities, providing in this way data for global comparisons.

In crisis situations in general and in a health crisis in particular, the media do play an important role in bringing societies in contact with risk (Waisbord 2002). During such situations, the traditional journalistic commitment to objectivity, which is the main source of their authority, is challenged (Kotisova 2020). “There is a heightened need to understand the nature of public trust in expertise, just as there is to puzzle through what it means to designate journalism as an ‘essential service’ or journalists as ‘essential workers’,” Lewis urges scholars (Lewis 2020, p. 686).

Therefore, this discussion becomes important for journalism to continue to play an important role in the public’s information about health issues, (a) as it can give insight that derives from the media content on how the journalistic profession distinguishes itself from alternative sources of information and (b) as it can provide information on journalistic authority, in other words, if “journalists still possess a right to create legitimate discursive knowledge about events in the world for others” (Carlson 2017, p. 13).

At the same time, it is worth looking at the differences between digital and print journalistic culture. The avalanche of human response is being facilitated by the flow of information from traditional media but, in particular, by the networked world of social media. Social media became a significant conduit for news and information in the modern media environment, with one in three people in the world engaging in social media and two-thirds of those on the internet using it (Ortiz-Ospina and Roser 2020). As stated by Islam et al. (2020) “since the onset of the COVID-19 pandemic, social media users have been playing a role in all stages of knowledge translation, including COVID-19 morbidity and mortality, interventions, spreading rumors and conspiracy theories, and reporting stigma”. Twitter has been shown to capture the dynamics of real-world events including the spread of diseases (Kagashe et al. 2017; Vijaykumar et al. 2018).

This research aimed at shedding light on whether online journalism can actually be functionally differentiated from print journalism, in a context like this, by using its technological component as a determining factor. Moreover, as the sources from which people obtain new information have been changing and social media such as Twitter have become an important tool for information, it is worth observing the discussion that took place on Twitter to reveal trends in the popular social medium. Though not the most popular social medium in Greece, Twitter was chosen as it is one of the more frequently analyzed and has become a prominent space for political talk. Twitter, as a text-based online service, permits researchers to recognize objects of political and public attention and compare them to the media agenda. Moreover, it was chosen as, since the outbreak of COVID-19, the number of tweets about the pandemic has been increasing (Singh et al. 2020, p. 20). It should be noted though that the specific sample aimed mainly at bringing forward some of the basic dimensions of the discussion rather than exhausting all possible aspects.

But first, we need to take a look at the environment within which all these take place and provide a brief background of mis/disinformation related to the COVID-19 pandemic.

1.1. The Neologism “Infodemic” as Point of Departure

In February 2020, the term “infodemic”, coined in 2003 by political scientist David Rothkopf (Rothkopf 2003), was first used in the context of COVID-19 by the World Health Organization to refer to the avalanche of false news, conspiracy theories, and misleading advertisements that flooded the news after the outbreak of COVID-19 (Muñoz-Sastre et al. 2021). The term “infodemic”, defined as “an overabundance of information—some accurate and some not—that makes it hard for people to find trustworthy sources and reliable guidance when they need it”, was coined to categorize some of the common features of rumors, stigma, and conspiracy theories during public health emergencies (Islam et al. 2020). “We are not just fighting an epidemic, we are fighting an epidemic of misinformation”, WHO Director-General Tedros Adhanom Ghebreyesus told the Munich Security Conference on 15 February 2020.

The “infodemic” neologism concerns essentially a large amount of information about a problem, which is an obstacle to its solution. It comes from the terms information and epidemic to outline the perils of mis/disinformation phenomena during the management of virus outbreaks. A major obstacle in such cases is the increase in the circulation of fake news and groundless allegations. Fake news is fabricated information that cleverly mimics real news and exploits existing public beliefs to influence or destabilize society and institutions, causing confusion and stress among citizens (Waisbord 2018).

Information is an important public good and at the same time a resource. In extremely complex situations, such as the coronavirus pandemic, it is both a tool for guidance and reassurance. In case of mismanagement, it spreads panic and chaos. When crises occur, such as wars, riots, terrorism, illnesses, and diseases, to name a few, the media gain an important role in mediating information (Lee and Bottomley 2010).

In this context, the media, by providing reliable and cross-referenced information, help citizens make informed decisions. By providing news, journalism becomes essential for the contemporary public sphere, which in liberal democracies functions as an independent and intermediate system between state and society and guarantees the principle of universal access to information for all citizens (Habermas 2006).

1.2. Media and Public Health Crises

During major crises, people rely more than usual on communication tools and the media (Lyu 2012). On a number of occasions, such as in Tai and Sun’s (2007) study of media dependency during the SARS epidemic crisis, Dalrymple et al.’s (2016) study on social media during the 2014 Ebola crisis, Holland et al.’s study on the legacy of the swine flu global pandemic (Holland et al. 2014), and Krishnatray and Gadekar’s study on the construction of death in H1N1 news in The Times of India, a body of research has examined the potentials and challenges of media usage in public health crises. The media play a particularly important role in raising awareness and informing the population so as to protect themselves from a virus, as in this case. In fact, the media have the potential to exert a great deal of influence for good or evil. In the recent pandemic of COVID-19, rumors soon circulated and affected the real world, often creating confusion and fear among internet users.

Important is also the role of expert sources. Most of the previous studies suggest that institutional actors and experts, in the more restrictive sense, account for the majority of cited sources. The role of experts in journalistic practice is considered central, and expert sources are used in journalistic texts primarily to provide facts, increase credibility, and ensure objectivity (Boyce 2006). They also shape the context of the story, are a dramaturgical component, and increase the news value by highlighting threat, creating proximity, or triggering controversy (Nölleke in Wagner et al. 2019).

The coronavirus pandemic was not the first to reach the age of internet connection and interdependence: at least three other global epidemics occurred in the previous ten years. The H1N1 pandemic, the Ebola epidemic, and the Zika epidemic have all had obvious and well-documented influences on social interactions and connections (Dalrymple et al. 2016; Ribeiro et al. 2018).

In the case of the new coronavirus, within a few weeks from its outbreak, a series of misleading rumors were circulated. Both the outbreak and the peak of the first phase of the pandemic coincided with the outbreak and peak of the infodemic. In fact, Depoux et al. (2020) reported that panic through social media traveled faster than the virus itself.

The COVID-19 “infodemic” brought to the surface misinformation in the form of unverified rumors, refashioned in the name of the coronavirus, that have always been there evident in controversial issues in the areas of health and science, such as climate-change denial, anti-vaccination, or anti-5G.

Some of the basic hoaxes that circulated were that:

• The virus is a biological weapon, developed in a lab.
• The vaccine exists and this pandemic is a “game” of the pharmaceutical companies.
• There are alternative therapies that can treat it effectively.
• 5G technologies are linked to the virus (Imhoff and Lamberty 2020, p. 1113).

1.3. The Coronavirus Pandemic in the Greek Media Ecosystem

The systematic coverage of the coronavirus pandemic by the Greek mass media began in February 2020, specifically, from the time the virus made its appearance in the most significant way in Italy. Until then and since January 2020, news about the virus had been sporadically visible depending mainly on the news reports of the international media and press agencies.

Greece belongs to a group of countries in Southern Europe that are characterized by distinct patterns of media—political interactions—and similar to Italy, Spain, and Portugal, the Greek media have developed within a system of clientelism (Touri and Kostarella 2017). In this system, specific political affiliations exist to which media content is tightly linked (Hallin and Mancini 2004; Papathanassopoulos 2013). These “eclectic affinities” have cost over the years mainstream media their credibility (Pleios 2013; Skamnakis 2018; Papathanassopoulos 2020). As the country in the past decade went through a deep recession, the media industry suffered major losses in advertising revenue and huge reductions in state subsidies, while public distrust increased (Papathanassopoulos 2013; Siapera et al. 2015).

As far as it concerns the “battle” between old and new media, as Papathanassopoulos (2020) notes, in contrast to most Western countries, Greek users tend to prefer or even trust pure digital news outlets more than those of the traditional mainstream media. Another surprising finding is that “Greece appears to be the only country in the world that believes social media do a better job in separating fact from fiction than traditional news media, but this is linked more to the low opinion Greeks have of news media in general rather than the quality of information provided” (Papathanassopoulos 2020, p. 139).

Despite the recession of the previous years, Greece has received a lot of praise from the international media as the country has managed the first stage of the COVID-19 crisis in an exemplary manner. The management included the decision by the Greek government to commit 11 million euros from the national budget to an urgent publicity campaign to promote measures to contain the spread of COVID-19. This raised criticism among opposition MPs and part of the press that the government is using the COVID-19 measures to improve its relations with the media without transparency and became an issue of dispute, evident especially in the Twitter debate.

2. Structure of the Research: The Combination of Two Data Sets

The main goal of this research was to uncover the characteristics of coverage in the Greek media through their print and online news editions. Specifically, through 7457 news items from 13 different media outlets, we tried to shed some light on key issues such as the type of journalistic content, the dominant themes, and the main protagonists and at the same time monitor the presence of hoaxes and conspiracy theories. At the same time, as social media is a significant conduit for news and information in the modern media ecosystem, we monitored the discussion on Twitter under the most popular hashtags (e.g., #COVID19 greece) in order to identify the dominant trends and compare them to the key findings of our media coverage. We took a first look at the amount of conversation taking place on Twitter, with respect to COVID-19, the themes of discussion, where the discussion is emerging from, myths shared about the virus, etc.

The main hypotheses of this research, informed by previous literature (Schudson 2003; Quandt 2008; Krishnatray and Gadekar 2014; Carlson 2017; Bernadas and Ilagan 2020; Falcone and Sapienza 2020; Boberg et al. 2020; Mellado et al. 2021) are as follows:

H1. 

The pandemic was covered by the Greek media in the light of an international problem, which is managed locally.

H2. 

COVID-19 represents a crisis context in which standard norms and practices may be significantly modified in different types of news outlets and social media platforms.

H3. 

The main sources of information in Greek media have been the officials and other “authorized knowers”.

H4. 

Mis/disinformation regarding COVID-19 has been abundant in the Greek media and on Twitter.

Our research methodology was content analysis with the aim of highlighting the patterns of coverage through their quantitative processing, using descriptive statistics (SPSS). Our content comes from two sets of data: the newspapers and news sites data set and the Twitter data set.

2.1. The Mass Media Data Set

2.1.1. Period of Study

Our newspapers and news sites sample comes from print and online publications starting from 13 March 2020. Three days before, on 10 March, Greek Health Minister Vassilis Kikilias announced the closure of all schools and universities as part of efforts to contain the spread of the novel coronavirus. Also, from 14 March, the Greek government has decided to shut down commercial stores, cinemas, bars, and restaurants, in response to the escalating coronavirus outbreak.

We stopped collecting data on 10 April 2020 when Greece managed to flatten the curve and optimism was evident as reflected in the statement of the spokesperson of the Ministry of Health, epidemiologist Sotiris Tsiodras (2020): “the epidemic curve of deaths has not increased, we just expect it to show a decrease in the coming days”.

During this period, media coverage was almost monothematic in all the publications under consideration, until April 10, when the de-escalation of the coverage became visible.

2.1.2. Sample

The sample (N = 7547) includes content from seven of the most popular sites in the country, excluding social media and banking and business group websites, as recorded by Alexa ranking (access date 3 December 2020). Also, it comes from the first four newspapers in terms of sales, as reflected in the bulletins of the distribution agency “Argos” (circulation bulletin, 13 March 2020), the newspaper Kathimerini, one of the most historical and important Greek publications (legacy press), but also the extreme views newspaper To Makelio, in order for the sample to be as representative as possible.

Newspapers sample: We analyzed all newspaper items that were relevant to COVID-19. Coronavirus was the keyword for the selection of the unit of analysis.

News sites sample: Given that a site can contain more than 100 articles per day, we applied the following selection rule: only the first twenty coronavirus articles on each site were coded, starting at 6 a.m.

Table 1. News sites ranking (according to Alexa ranking 3 December 2020).

Rank	News Site
5th place	In.gr
6th place	Lifo.gr
8th place	Enimerotiko.gr
10th place	Zougla.gr
18th place	Protothema.gr
29th place	Iefimerida.gr
37th place	Newsbomb.gr

and

Table 2. Newspaper circulation (13 March 2020).

Newspaper	Circulation Numbers
Ta Nea	11.980
I Efimerida twn Syntaktwn	5.580
Eleftheros Typos	3.760
Eleftheri Wra	2.670
Kathimerini	No data available
Makelio	No data available

Source: Argos Distribution Agency.

are presented below for a better understanding of our sample, while Figure 1 is a snapshot of the newspapers on 13 March 2020.

2.1.3. Unit of Analysis and Coding

The unit of analysis is the news item, in other words, each autonomous information section on the sites and newspapers under examination. The analysis of the news items was based on a codebook, focusing on the content categories, presented below, under study. The final form was prepared on the basis of the previous pilot form and expert comments on it. The sample was coded manually.

Our sample (N = 7547) was analyzed according to:

• The type of the news item (article, interview, op-ed, etc.) (related to H2);
• The presence of images (related to H2);
• Whether they are signed by an author or the editorial team (related to H2);
• The sources from which the information mentioned in the text comes, e.g., experts, global organizations (e.g., WHO), people who got sick (related to H3);
• Whether it focuses on human drama or data (related to H2);
• It contains hoaxes and conspiracy theories, e.g., the virus was developed in the lab, etc. (related to H4);
• The type of content (related to H2);
• The main source of reference (related to H2 and H3);
• The presence of a link (coding only applies to sites) (related to H2 and H3);
• The main theme (related to H1).

Twelve annotators were involved in the process of coding, all of them trained by the principal investigator(s). At this point, it should be underlined that one of the limitations of this research is the fact that the inter-annotator agreement was not verified.

2.2. The Social Media Data Set

Besides analyzing the news items coming from print and online news media which reflect the tendencies of the public debate as expressed by opinion leaders, we chose to analyze tweets about the same event (COVID-19) but at a different time period, 29 May–14 June 2020, which coincides with the end of the lockdown. This choice was made in order to understand the structure and the characteristics of the public debate, as represented on Twitter.

This part of our research looks at the amount of conversation taking place on Twitter, with respect to COVID-19, which were the main sources that discussion was emerging from, and the main themes of discussion.

Using the Twitter Streaming API, we began collecting tweets (N = 36,317) related to COVID-19 from 29 May to 14 June 2020. Twitter’s streaming API returns any tweet containing the keyword in the text of the tweet. For example, the keyword “COVID” will return tweets that contain both “COVID19” and “COVID-19” or “COVID19GREECE”.

Table 3. Hashtags and date of collection.

Hashtag	Tracked Since
#COVID19 greece	29 May 2020
#μενουμε_ασφαλεις (stay safe)	29 May 2020
#κορονοιος (coronavirus)	29 May 2020
#πανδημια (pandemic)	29 May 2020
#COVID_19	29 May 2020
#καραντίνα (quarantine)	29 May 2020
#μμε_ξεφτίλες (media are a disgrace)	29 May 2020
#Τσιοδρας (Tsiodras: the name of the person in charge of Greece’s management of the coronavirus pandemic)	29 May 2020

shows the hashtags we used to collect data. By continuously monitoring Twitter’s trending topics, keywords, and sources associated with COVID-19, we did our best to capture conversations related to the coronavirus pandemic for the specific period of time. During the study period, the overall number of tweets was 36,317, quotes 3607, and retweets 23,337.

Figure 2 shows the overall volume of tweets and the number of new COVID-19 cases for the same period. The event that propelled the rise in the number of tweets between the 5th and the 8th of June is the massive traveling of Greeks for the three-day weekend of the Orthodox Holy Spirit holiday.

The limitations to our data set cannot remain unmentioned. The first is that we collected our data set leveraging Twitter’s free streaming API, which only returns 1% of the total Twitter volume, and the volume of tweets we collected was dependent on our network connection. That resulted in a gap in our data set due to intermittent internet connectivity issues on the 6th and 7th of May.

3. Findings

In this section, we present the main findings derived from our research. It is divided into two parts: the first is about the media content as it appeared in press and news sites, while the second is about the content of conversation on Twitter.

News media content is analyzed according to some of the formal characteristics, such as the type of news item and the type of content. Then, we continue with the main protagonists and themes covered, and at the end, we discuss the presence or not of hoaxes and conspiracy theories.

As for the content of conversation on Twitter, it is analyzed through its main themes of discussion, trending topics, and the main sources (social media accounts) it derives from.

3.1. Media Content

Figure 3 shows that the report prevails in print and online media. Besides reports, there are a fair number of stories and subjective forms of writing. In the second place for online and in the third place for print media comes the story. Under the category “other”—to a greater extent in print media and to a lesser extent in online news media—we mainly come across interviews and secondarily other types of news content, such as letters to the editor, commentaries, etc.

3.2. Source of Information and Use of Hyperlinks

In our research, we coded any explicit reference to a source at the beginning or end of the text but also any tacit assumption of the original source of information (with reference within the text), following the methodology of Quandt (2008).

Earlier research indicates that online journalism is highly dependent on external sources and parent media. The “copy and paste” principle seems to be acceptable for some online journalists; therefore, “shovelware” is the core material for many of the smaller news sites (Quandt 2008, pp. 727–28).

Our findings show that this does not only apply to online journalists but also to print journalists, since to a very large extent, the original source of information is not identified.

As shown in the relevant figure (see Figure 4), a difference between print and online media is observed, as print media contain a significantly higher percentage of original reporting through primary investigation as opposed to online media which prefer curating content coming from other sources. A possible explanation is that prestige press still puts original reporting at the forefront. In Greece, as Skamnakis (2018) points out, nine out of ten internet media sites republish articles that are mainly produced in print media without the permission of their authors, who are in the majority professional journalists. Our finding that print media do not rely heavily on agency material, despite the fact that news agencies have for many decades been considered the main providers of raw journalistic content, is consistent with Quandt’s study (2008) (Baerns, cited in Quandt 2008).

At this point, it is worth adding to our discussion the term “churnalism”, which is used for a form of journalism that is a result of the growing need for news content (Saltzis 2012) in which press releases and other forms of prefabricated material are used to produce journalistic content in newspapers and other media, a “passive processing of news material”, as Davies puts it (Davies 2008, p. 59). The term moved into mainstream journalism discourse, describing in less-than-complimentary terms the recycling process of news production which drew increasingly on wire service copy and public relations (PR) subsidies (Johnston and Forde 2017, p. 943). Saridou et al. in 2017 through quantitative analysis provided empirical evidence that the recycling of news content from established elite sources and across popular news sites has increased between 2013 and 2016, and their main outcome was that “the current practice of churning designates that established elite sources and actors are the main shapers of the agenda and viewpoints circulating on the Web” (Saridou et al. 2017).

It is possible, therefore, that many print and online media do not follow a transparent and systematic editorial process, while the high degree of presence of news items in which the original source/author is not identified testifies to the growing risk of copy–paste becoming a common practice for everyone.

Regarding the presence of hyperlinks, it should be mentioned that 62.1% of all online publications include a hyperlink (link). The hyperlink mainly leads to an older post on the same news site and more rarely to another site. Thus, one of the basic functions of multimedia journalism is not utilized properly.

In terms of basic actors (see Figure 5), top-down information from the Greek authorities is dominant. Our findings regarding the main protagonists are fully consistent with what has been recorded so far about the dominance of authorities regarding health crises in the media agenda. Shih et al. (2011, pp. 17–18) explain this dominance of institutional and administrative sources not only by the nature of health issues but also by the fact that credible and authoritative sources meet the professional requirements and routines of journalists. As she notices, journalists tend to present health issues from a rather official and general than from an individual human point of view.

In most of the publications, the government, through the daily press conference of the spokesperson of the Ministry of Health, Professor Sotiris Tsiodras, and the Deputy Minister of Civil Protection, Nikos Hardalias, was present as a key player in the media arena. This is a reasonable finding given the fact that especially during the first period of the general lockdown (March–April 2020), when fear prevailed, the main source of information was the press conference held at 6 o’clock in the evening by Tsiodras and Hardalias on new coronavirus developments. The professor and infectious diseases expert became a household name due to the daily briefings he held for 72 consecutive days.

What is interesting to note is that the “rally-round-the-flag” effect was observed worldwide, while it was accompanied by an increase in the popularity of political leaders, which may not have had anything to do with their actual performance in crisis management, as Will Jennings, Professor of Political Science and Public Policy at the University of Southampton, noted on March 30 (The UK in a changing Europe, 30 March 2020).

The “rally-round-the-flag” effect coined by John Mueller (Mueller 1970) suggests that in times of crisis when the nation is threatened, incumbent leaders benefit from a rise in public support. It is a concept used in political science and international relations to explain the increased short-term popular support of a country’s government or political leaders at times of crisis or war. Mueller (1970) defined it as coming from an event with three qualities: (a) has global range, (b) directly involves the country and its leader, and (c) is specific, dramatic, and intensely focused. Based on the above, it can be considered that the COVID-19 pandemic shares these characteristics.

The experts come in the second place in the online media coverage and in the fourth in print. This finding is compatible with previous research, as when it comes to complex topics in the field of health and risk communication, as aforementioned, experts are of high importance for the credibility of a news media report, and they are regarded as particularly influential sources (Page et al. 1987, p. 39).

Celebrities, and other prominent public figures, come second to last both in online and print media. “The pandemic has disrupted relations among the masses, the elites and the celebrities who liaise between them,” Amanda Hess (Hess 2020) wrote in the New York Times on 30 March 2020. In the last place in both print and online media come politicians either as individuals or through their parties.

3.3. Topics

The prevalent topic categories in both print and online media, as we can see in Figure 6, are international and society–human interest stories but in a different order.

These findings are consistent with the fact that in the outbreak of the coronavirus, the coverage of the pandemic in neighboring Italy was at the top of the agenda on a daily basis, while in the days that followed, an important part of the agenda was devoted to the pandemic management in the European countries. The trend for a transnational approach is compatible with the findings from Finland and the European Journalism Observatory, where it is pointed out that “since COVID-19 spread in Italy, news coverage in Finland has been explicitly transnational. Instead of focusing primarily on the impact that the virus may have on the Finnish population, news and commentary have focused on other countries” (Hekkilla 2020).

The topic category “society–human interest stories” includes a wide range of issues related to how the population will follow the measures to prevent coronavirus, how the churches will operate during Easter time, and also the measures which were to be put in effect during the Easter period. In the third place, the topic category “health and science” hosts the latest developments on antibody tests, clinical trials of vaccines, etc.

3.4. Hoaxes and Conspiracy Theories

As earlier mentioned, the term infodemic has been coined to describe the dangers of misinformation phenomena during the management of virus outbreaks, since it could even speed up the epidemic process by influencing and fragmenting social response (Cinelli et al. 2020).

While social media platforms, such as Twitter, may amplify rumors and questionable information, our assumption, regarding print and online media is that professionalism can function as a retarder in the dissemination of such narratives.

As we observe in the relevant figure (see Figure 7), the systematic presence of hoaxes and conspiracy theories in print and online media is not confirmed. Conspiracy theories are almost entirely absent. Their presence is very limited, with the exception of the newspaper Makelio which displays a significant number of articles containing references to unsubstantiated rumors and the newspaper Eleftheri Ora to a lesser extent. Both Makelio and Eleftheri Ora are rather controversial newspapers. All the other media take a critical view. An enlightening example of the attitude of prestige press toward conspiracy theories is the article in Kathimerini entitled “Looking for the vaccine for the COVID 10 conspiracy theories” (“Kathimerini”, 5 April 2020, p. 13).

3.5. Human or Data Stories?

While data journalism has been developing in recent years worldwide, the Greek media have not allocated special resources, material and human, for its development. The coverage of the coronavirus crisis gave a boost to data in the form of infographics and statistics. Personal opinions and human stories, as told by their protagonists, were also part of the media content.

By examining the content of online media, we seek to highlight the extent to which they make use of their multimedia capabilities, which have been mentioned before in relation to hyperlinking. As we can see in the relevant figure (see Figure 8), most of the online media use text as their main type of content, and only one-third offer a combination of multimedia content. Statistically insignificant is the presence of news items that contain only video or audio.

4. Discussion on the Content of Conversation on Twitter

In this section, we take a look at the content of the conversation taking place on Twitter. By continuously monitoring Twitter’s trending topics, keywords, and sources associated with COVID-19, we did our best to capture conversations related to the pandemic and in particular the end of its first phase, known as the lockdown period (karantina). What are the most prevalent words and themes of discussion taking place about COVID-19?

As we see in the word cloud (Figure 9) and the relevant

Table 4. Top 10 most frequently used words in tweets.

Word	Volume
Quarantine	14,963
Pandemic	7206
Mitsotakis	272
Lockdown	245
Italy	186
Vaccine	139
Theodoridou	105
Tsiodras	110
Chardalias	90
China	66
Trump	65

, the most frequently used words in tweets are quarantine and pandemic.

Table 5. Most popular accounts by number of retweets.

Account (Screen Name)	Number of Retweets
@MDenaxa (journalist)	527
@paganiotis (unknown id)	423
@serkot65 (Serafeim Kotrotsos, journalist)	301
@PithikosTapas (unknown id)	242
@Kinima_Ypervasi (self-identified as public movement)	208
@ManosVoularinos (comedian)	219
@Teletai_Mpouk (funeral services)	207
@Angelakard (unknown id)	145
@Kanekos69 (unknown id)	132
@GltLucia (unknown id)	122
@bkex (Voula Kexagia) journalist)	122

shows the most popular accounts in our sample. It is worth noticing that the accounts of two journalists are in the first three places. The finding proves that journalists are considered an important source of information. On the other hand, as we see in

Table 6. Official media accounts by number of retweets.

Media Account	Number of Retweets
amna_news (Athens News Agency)	115
Iefimerida	89
Protothema.gr	57
Makeleio.gr	55
Ethnosgr	24
Kathimerini_gr	22
zougla_online	20
In.gr	10
Newsbomb	9
ta_nea	6

, the official media accounts are not popular, as their tweets are very few times retweeted.

Main Topics

From the analyzed tweets, we identified topics that were grouped into seven themes: the fear of a second lockdown, the violation of the quarantine rules and the specific case of the daughter of a famous Greek singer that received massive attention, issues of public health, xenophobic and racism issues, conspiracy theories, criticism of the government, and humorous tweets.

Theme 1: The fear of a second quarantine

The topic with the most retweets was the possibility of a second quarantine with 1112 references.

Examples of the most retweeted are presented in

Table 7. Tweets regarding the fear of a second quarantine.

Account (Screen Name)	Tweet	Number of Retweets
@a_olimpia	If we go for a second lockdown, instead of a bathing suit we will need a gaberdine	45
@variemaikpeinaw	Sweet boys and girls put your masks on until I come so the second lockdown finds me in Greece	22

:

Theme 2: The violation of the quarantine rules and the case of the young student that returned from the UK

Another one of the most popular topics for the same period has been the party held by the daughter of a famous Greek singer who returned from her studies in the UK and did not comply with the rules for a 14-day quarantine.

Most of the tweets, on the day news about the party was spread (2/6), have been about that party. A total of 344 tweets contain the term “student” and 105 the term “Theodoridou”, who is the mother of the girl. The most popular of the tweets has been the one that used the title of a famous song by the mother of the student “Wanted” (Kataziteitai) to refer to the issue.

Examples are presented in

Table 8. Tweets regarding the violation of the quarantine rules.

Account (Screen Name)	Tweet	Number of Retweets
@paganiotis	The student who broke the quarantine and held a party is wanted	94
@PithikosTapas	Wanted, wanted (after the song of her famous mother)	101
@ManosVoularinos	I think we all know that someone who returns from the UK and breaks the quarantine should pay a fine	10
@gesta_res	The daughter of a well known family in Thessaloniki returns from the UK where she studies and doesn’t follow the quarantine as obliged	3
@IsidorosLos	Yes, mate. The daughter of Theodoridou and businessman Mpetas came back from London and didn’t stay in quarantine for 14 days	9
@Sampsonius_	The girl returned via Bulgaria. She was so happy that broke the quarantine	18

:

Theme 3: Public health

The issues of public health keep special visibility. One of the most retweeted posts has been that by @GltLucia: “If the pandemic has taught us anything it is that Public Health with its human resources and structures is irreplaceable”. On 29 May, it was the most popular tweet with over 120 retweets.

On the same day, the tweet by journalist Serafeim Kotrotsos (@serkot65) brought to attention the case of a top pediatric heart surgeon, who asked to volunteer during the pandemic at “Agia Sophia” children’s hospital and the administration of the hospital reportedly refused to let him. This was the most popular tweet with 290 retweets:@serkot65: “Top pediatric heart surgeon asked to volunteer for the pandemic and the administration of “Agia Sofia” provocatively ignored him.”At the same time, RT @krissgr expressed criticism of the government’s handling of public health issues, stating that “public health is “the best part of the meat” which should be given to the owners of private clinics” (22 retweets).

Theme 4: Racism and xenophobic issues

A special place on the agenda was held by issues of migration and refugees. Criticism is being expressed that refugees are better treated at the expense of Greek citizens.

Here are some tweets of this category:

@george_lykos: “Look here at what happens in Pedio Areos (park in Athens where refugees gather). All women wear scarves and carry prams with children and we offer them everything for free.”

Another one with 140 retweets is that by @angie22gr who writes “the illegal immigrants (referred as lathro) smashed everything in a hotel in Ermionida because they don’t like the quarantine. My little birds are oppressed. Whoever comes to escape from any hell and is hosted with all his good deeds does not react like that. Should they be deported right now?”

However, at the same time, a different opinion is expressed by @FakeVolante who says, “we pay the quarantine for the all-inclusive tourists and we leave the children refugees under the rain and the hot sun” (47 retweets).

Theme 5: Conspiracy theories

Since the onset of the COVID-19 outbreak, several conspiracy theories had been circulating. None of them appeared in our data, except for a few negative references to Sotiris Tsiodras, the doctor in charge of Greece’s management of the coronavirus. These very few tweets say that he is sponsored by the big pharmaceutical company GlaxoSmithKline and that he is the one who suggested the purchase of the vaccines for H1N1 which went to the garbage a few years ago.

The tweet by @gelosoil calls him a vaccine seller, a crook. At the same time, tweets that question the existence of COVID-19 are scarce, such as this one addressed to the Prime Minister:

@PrimeministerGR: “When you say scientific community, you mean the crook, the thief Tsiodras who presented false information about deaths and did not know what a pandemic is and if a mask is needed.”

Theme 6: Criticism of the government

The most popular is the tweet by journalist Maria Denaxa (@mdenaxa) who questions the decision of the government to cover the costs for the tourists who come from abroad and have to stay in quarantine: “The Greek tax payer will pay 135 euros per day and per traveler for the quarantine of the tourists that come to Greece”. This was retweeted 527 times.

As we can observe in

Table 9. Tweets conveying government criticism.

Account (Screen Name)	Tweet	Number of Retweets
@tetRadio	Efimerida twn Syntaktwn (daily newspaper) received 39 thousand euros for the campaign # stay safe	64
@el_ges	The most anti-educational bill, concerning the future of our children, was passed by #New Democracy kseftiles (a word used to show disgrace) and at the same time money were given to non-existent media.	6
@ofp8791	@SteliosPetsas (government spokesman) said that #Documento didn’t get advertising money because it says that #COVID19greece is a lie. What a punk!	4
@_plystra	I live in a country that frees pedophiles, closes hospitals that treat children but advertises the campaign #stay safe	22
@aigeas	The 20 million “for our protection” from the coronavirus were distribute to “ghost” companies and to the pockets “of our favorites”.	10

the most serious field of controversy between the government and the opposition that was transferred to Τwitter has been the distribution of money in each medium for the campaign #stay safe (menoyme asfaleis):

Theme 7: Humorous tweets

Finally, there is the category of humorous tweets. In this category, a funeral home posts the most popular tweets. It is the account @Teletai_Mpouk with 208 retweets.

Another tweet indicative of this group of tweets is that by @Boubounokefalos: “It is not that I got fat during the quarantine, but now I have to use air freshener” (instead of perfume). This was retweeted 172 times on 31 May 2020.

5. Conclusions

This research was an attempt to answer some initial questions regarding the modus operandi of media and digital platforms in the COVID-19 era, most importantly, their contribution to the rise/decline of the authority of scientific expertise and their role in mis/disinformation around health and science controversies produced, distributed, and redistributed in offline and digital environments. It also tried to reveal different patterns in journalistic conventions between old/new media and social media, using the Greek information ecosystem as a case study.

The COVID-19 pandemic due to its unprecedented dimension and intensity, but also its disruptive nature, is an important opportunity for study that will provide additional tools for enabling journalists to help the public make better-informed decisions in similar situations in the future. At the same time, it gave us the opportunity to study the role of professional media outlets and social media in the circulation of information and their contribution to helping people make better sense of the pandemic.

According to our findings, the Greek media faced the crisis “with a view to the world”, emphasizing international coverage, giving priority to the authorities and scientists, and keeping (at least in their majority) hoaxes and conspiracy theories out of the agenda. On the other hand, the timeline of the Greek Twitter is related to current events, either as updates to an unfolding story or as reactions to ongoing events, a finding consistent with the observation by Hermida and Mellado (2020).

What we observed in our research is that:

#1: News coverage in Greece has been explicitly transnational, focusing on other countries and especially the Mediterranean ones, which have been hit the hardest by COVID-19.

#2: Different patterns in journalistic conventions are observed between our two data sets, especially regarding the rhetorical practices. Both print and electronic news media seem to follow traditional patterns of news reporting. Tweets tend to be more emotional and, in some occasions, very different to the rhetorical practices of traditional journalism as the verbal and visual style of the platforms favors short and catchy messages, while retweets are means to assign prominence and visibility (Hermida and Mellado 2020). Also, the use of Twitter by journalists is very interesting, as from what our data set reveals, some of them are very popular and active on the Greek Twittersphere. It would be an interesting future work to examine the tweeting norms and habits (self-branding, personalization) of professionals. For example, during public health crises, do Greek journalists use Twitter as a tool for gathering and disseminating news or as a vehicle to express criticism? An initial observation is that the Greek Twittersphere is not so much a significant source of events whose rapid development cannot be captured live by traditional platforms but rather an opinion forum, where diverse and often critical views are expressed.

#3: The authorities set the tone: our findings regarding the basic actors are fully compatible with what has been recorded about the dominance, globally, of the authorities in the media agenda. In most of the news items, the government, through the daily press conference, was present as a key player in the media arena. The response to the government’s actions—at least at the first stage of the COVID-19 outbreak—was almost entirely positive, verifying the “rally-round-the-flag” phenomenon, which suggests that in times of crisis when the nation is threatened, incumbent leaders benefit from a rise in public support. On the other hand, experts and the representatives of scientific institutions are given prominence in media coverage, and therefore, the spreading of misinformation is limited.

#4: Extensive presence of hoaxes and conspiracy theories in the print and electronic press is not confirmed. We should also underline that we encountered only a limited number of tweets circulating fully fabricated stories. That is interesting as there is a widespread belief that for social media, unlike more traditional outlets, the spread of information is less tied to content accuracy. Consistent with Brennen et al. (2020, p. 6) is also our finding that across the sample, the most common claims within pieces of misinformation concern the actions or policies that public authorities are taking to address COVID-19, whether individual national/regional/local governments, health authorities, or international bodies such as the WHO and the UN. Therefore, we propose that we need more evidence about the types and sources of COVID-19 mis/disinformation to reach conclusive results.

This study does not pretend to capture all aspects of the problem but rather brings forward some of the basic parameters that can enlighten the complex relationship between different platforms in the context of the COVID-19 pandemic.