Abstract
Purpose
Fact-checking has been changing in recent years from an initial stage in which fact-checkers were more concerned with political discourse to a stage in which combating misinformation becomes the primary purpose. This work examines more closely the standardizing and the customizing aspects of active fact-checking outlets in Portuguese-speaking countries, focusing on the verification methods and organizational models in use.
Design/methodology/approach
Based on Content Analysis, we collected manually 318 posts during June 2019 from each fact-checking outlets website and then examined each post according to six general concepts: discourse, sources, context, classification, graphic representation, and financing. There were 15 active fact-checking outlets in Brazil (13) and Portugal (2). No active outlets were found in the African countries.
Findings
Although there is room for inventiveness in fact-checking practices, it is restricted to the classification models adopted and the graphic representation demanded by them. Only two largest Portuguese-speaking countries (Brazil and Portugal) have active fact-checking initiatives during the study period. In Mozambique, we found the outlet named Mozcheck that was inactive with no published content during the research period. From our analysis, we detected a pattern between the type of misinformation and the media to which it is most often linked: false information was circulated mainly in texts, while false contexts were mainly circulated in videos and images led to more manipulated content. In addition, in relation to the sources used to verification of the content, we noticed a large volume of posts relied only on sources came from contacts with press offices – this was especially true for political issues.
Practical implications
The analyzed data indicates that the standardization tendencies are related to the connection of these initiatives with traditional media. While the contrasting aspects of the fact-checking practices are related to the classification models and the graphic representation created by the outlets.
Social implications
It indicates that fact-checking outlets is still tied to traditional media in terms of its organizational and institutional business model. Inventiveness and innovation are restricted to the practice of fact-checking conducted by journalists and other professionals.
Originality/value
This is the first study to compare the practice of fact-checking in Portuguese-speaking countries and, besides looking at aspects of journalistic practice, it also seeks to analyze organizational elements of fact-checking outlets.
1 Introduction
There are two main reasons why fact-checking is performed: to verify political and public statements (Amazeen 2013; Graves 2017; Wintersieck 2015, 2017) and to fight disinformation (Walter et al. 2019; Young et al. 2018). Obviously, these two purposes do not fully embrace the practice of journalistic verification of facts and data publicly exposed under this modality (Brandtzæg and Følstad 2017; Lowrey 2017). In addition, they do not encapsulate all the advantages this practice has for the quality of information, and do not substantiate its efficiency when facing these challenges (Amazeen 2015; Clayton et al. 2019; Uscinski and Butler 2013; Walter et al. 2019; York et al. 2019; Young et al. 2018).
Studies have shown the growing relevance that fact-checking initiatives have attained over recent years in verifying political discourse (Graves et al. 2016; Nyhan and Reifler 2010, 2015; Spivak 2010; Wintersieck 2017). This is especially true during electoral periods when the amount of information searched on fact-checking organization webpages increases exponentially and, as a consequence, so does the rate of access (Wintersieck 2015). Furthermore, the influence that verifications have on voters indicates that a candidate’s performance in an electoral debate is verified as positive if the veracity of their statements is confirmed, and negative if they are verified as dishonest (Wintersieck 2017). Another finding is that people are much more likely to vote for a candidate if the fact-checks indicate that a particular candidate is honest (Wintersieck 2017).
While fact-checking political discourses measures the level of influence these discourses have on the public’s perception of political events; the verification of disinformation measures its effectiveness in combating fake news. Walter et al. (2019) identify the positive effects of checking for misinformation at the political level, but warn that this becomes less effective when using true scales, showing only parts of a claim or statement given in a campaign to be false. Moreover, the ability to correct political misinformation through the verification of facts is hindered by readers’ pre-existing beliefs, ideologies and knowledge.
The studies mentioned above show that, in addition to verifying information or statements, the formats and the use of visual elements for verifying are key to the impact and efficiency of fact-checking. So, in a context where formats and methods matter, how are the Portuguese language checking outlets structured to deal with the dissemination of false statements and news? Therefore, we address the following research questions.
About the fact-checking outlets,
RQ 1 – What is the connection between fact-checking initiatives and the organizational models of the other media outlets in each Portuguese-speaking country?
RQ 2 – What are the funding strategies of the fact-checking outlets?
About their publications,
RQ 3 – What discursive elements structure the verification methods of fact-checking outlets in Portuguese-speaking countries?
This article examines more closely the standardizing and the customizing aspects of current fact-checking initiatives in Portuguese-speaking countries, focusing on the verification methods and organizational models in use. The analysis of this fact-checking initiatives seeks to offer a contribution to studies in the area, particularly when we consider the absence of literature on fact checking in Portuguese and the predominance of Anglophone in comparative studies on the topic. The present study offers a contribution to multilingual perspectives in research on the most diverse phenomena, especially in the field of media studies and communication studies. We argue that, although there is room for inventiveness in verification practices, it is restricted to the classification models adopted and the graphic representation demanded by them.
2 Fact-checking practice, between the traditional and the innovative
In summary, the journalistic practice of fact-checking consists of “a systematic publication of assumptions on the validity of statements made by officials and public institutions, with the explicit intention of identifying whether a statement is factual” (Walter et al. 2019, p. 2). Over the last few years, in addition to evaluating the accuracy of statements made by public entities, fact-checkers have been dedicated to combating disinformation, and are then responsible for certifying the veracity of information distributed on websites, social media and messaging software.
Graves et al. (2016) indicate that the initial spread of fact-checking outlets within the United States is driven more by their appeal to professional values and status concerns on the part of journalists in political coverage than by indicators of public interest. However, the emergence of fake news as a global problem (Tandoc et al. 2020) seems to have given the fact-checking practice a potential leading role in validating information to a wide audience that is eager for qualified coverage in a broad thematic scope (Bode and Vraga 2018; Tandoc et al. 2018a, 2018b). While the growth of verification initiatives in the early years is related to journalists’ concern with the constitutive values of the profession, as indicated by Graves et al. (2016), on the other hand, the limited response of the public and the funding restrictions of these initiatives may have directed the actions of the verification agencies in recent years to combat disinformation.
In observing the development of the fact-checking initiatives, we perceive two important dichotomies: one related to the traditional and at the same time innovative aspects of the models elaborated by the fact-checking outlets, and the other related to the relationship between their global aspects of standardization and the singularity of their practices. These dichotomies appear in a diffuse way in studies published so far, despite being problematized in much of the literature. It is from these dichotomies that we seek to understand the fact-checking practices carried out by Portuguese speaking initiatives.
Although its emergence goes back thirty years, with initiatives to verify the political discourse on television (Cappella and Jamieson 1994; Dobbs 2012; Lowrey 2017; Young et al. 2018), it was with the expansion of access to information and to production of content, besides the consequent pluriferation of sources that led the verification of facts to its current form (Graves 2013; Lowrey 2017; Wintersieck 2015). Verification is a primordial stage of journalistic work and, in this sense, a consolidated activity, and even little reflected in professional practices. On the other hand, in this new informational context, the verification activity is repositioned as a novelty, even as an innovation – in the sense of in the sense indicated by Garcia-Avilés et al. (2018, p. 27), “the capacity to react to changes in products, processes and services through the use of creative skills that allow a problem or need to be identified, and to solve it through a solution that results in the introduction of something new that adds value to customers and to the media organization”.
In this context, the fact-checking initiatives offer creative solutions to the news verification models as they also resemble pre-existing models used by established media companies. These models, which are a mix between the new and uncertain and the pre-established and stable, contribute to the growing legitimacy of fact-checking initiatives (Lowrey 2017) and to their consolidation in the media field. Lowrey’s findings indicate that.
(1) media entities seek legitimacy within the context of other institutions; (2) media entities tend to want to look like one another; (3) common best practices evolve and shape decisions in a field; (4) media entities may decouple or buffer the effects of external change; and (5) the cumulative drag from the persistence of older media entities shapes the field (Lowrey 2017, pp. 390–391).
In addition to the dichotomy between traditional and innovative aspects of fact-checking models practiced around the world, we also found discussions around standardization and uniqueness. Standardization processes implicate the development and promotion of transorganizational legitimized models, seen in the recognition of fact-checking as a journalistic practice in different organizational contexts, but also from the normalization of policies and work methodologies, the definition of transorganizational rules and adoption of practices considered successful in previous experiences. As Graves (2018) indicates, despite presenting divergent practices, missions and organizational models, verification initiatives share broader values such as promoting democratic discourse and responsible government. The practice of validating discourse and facts connects with the ideals of professional journalism, such as precision and justice. However, Graves (2018) points out that these same ideals can be practiced in different ways.
For example, the ethical principles of accuracy, independence, collaboration, justice and transparency (“Ethics Policy – Poynter” n.d.) are the foundation of fact-checking journalistic practice according to The International Fact-Checking Network (IFCN). These principles guide the initiatives that are part of the IFCN, configuring verification methods and organizational and financing models created worldwide. But how these values are reproduced in fact-checking practices, whether or not belonging to this network, remains a relevant question.
From this perspective, it is also possible to see the development of the fact-checking journalistic practice through the models and/or methods proposed by the checking outlets. In this sense, there are elements that indicate the standardization of the models at the same time that a singularization is desired. In some cases, such as those of the initiatives that are part of IFCN, we perceive more clearly, in general terms, the steps that constitute the verification methods: (a) the choice of the object of validation (speech, data, statement, facts, hoaxes and events), (b) the contact with the sources to corroborate the information to be verified, (c) the reconstruction of the context in which these objects of verification became public and them are confronted with the information gathered from the sources, (d) the classification according to the verification labels of the fact-checking initiatives and (e) the final graphic representation.
Still there is a wide range of methods or models being practiced by the verification outlets around the globe (Graves 2013; Graves and Cherubini 2016; Vizoso et al. 2019), each one with its own particular features, which in turn gives the initiatives their uniqueness. A more accurate look at the diverse practices of fact-checking should contribute not only to show elements of differentiation between the initiatives, but also to provide valuable insights on the reproduction of journalistic values in contemporary times.
3 Gatekeeping theory and fact-checking
Statements, evidence and opinion. The verification of a piece of disinformation or a political discourse practiced by fact-checking outlets begins with the selection of the material to be verified, goes through the consultation of sources to corroborate the information, advances with the editing of the material and ends with the publication of the note in accordance with the media’s communication strategies. This whole process is subject to the decisions made by the fact-checkers as they go through the search for the veracity of the information. This example reflects the most traditional dimension of the gatekeeping theory approach in communication and journalism, which is to understand the different types of power that condition the making and publishing of news.
In communication and journalism studies, gatekeeping theory, in general terms, studies the process by which information is controlled as it passes through a filter. This filter is associated with the imposition and/or influence of different types of power in the selection of news and other media publications. In short, it is the theoretical perspective that seeks to understand why some news is published while others are not.
Gatekeeping also encompasses other dimensions of journalistic practice beyond understanding journalists’ decisions. Shoemaker and Vos (2009) argue that the theory also focuses on the levels of organization, social institution and social system.
It is from these levels or dimensions that Barzilai-Nahon (2011) considers gatekeeping the result of influences that aim to control the flow of information. Barzilai-Nahon (2011) also points out that these influences come from the internal policies of organizations, governments and elements of the social structure such as ideology and culture.
Therefore, in communication and journalism studies, research under the gatekeeping framework covers the individual factors of journalists, the characteristics of the information published by the media, the routines and internal organizational policies of the media, the political, cultural and economic pressure of the social and institutional context in which the organizations are inserted. In this text we focus on the influences of the social and institutional environment of the media.
Although the study explores aspects of the publications of fact-checking organizations, in this text we approach gatekeeping from the levels of organization, social institution and social system, according to Shoemaker and Vos (2009). That is, in this case, “we consider the result of influences that aim to control the flow of information” (Fakida 2021, p. 128). Our focus is not the decision process of the journalist fact-checker and the internal policies of these organizations, but the conjunctural elements of social, political and cultural orders that exert influence on this recent journalistic organizational model.
4 Methodology
Data collection for this research assessed fact-checking initiatives in Portuguese language countries. Our analysis corpus is comprised of the nine member states that make up the Community of Portuguese Language Countries (CPLP): Angola, Brazil, Cape Verde, Guinea-Bissau, Equatorial Guinea, Mozambique, Portugal, São Tomé and Príncipe, and East Timor.
We selected media outlets which have an online presence and looked at the active fact-checking initiatives in each country. Data collection took place between March and June 2019, considering data availability. The period also observed the regularity of journalistic coverage, since, at the time of collection, there were no discrepant events in any of the countries that were part of the analysis. In this period, the research was divided into two stages: (1) searching the web via Google for fact-checking outlets in the mentioned countries and monitoring, during March, April and May; and (2) collecting the verifications published by the fact-checking outlets on their web pages, during June. In the first stage, we identified outlets that, although at a certain period in the past had a constant and periodic publication regime, were not publishing new verifications during the months of monitoring. These initiatives were considered inactive and excluded from the second stage of the research. This was the case with Mozcheck, for example. Data collection was performed manually, with the support of a structured table based on the analysis criteria, as described in the codebook that integrates the research.
For the data analysis, we used the general principles of Content Analysis (Bardin 2011; Bauer 2015) and fact-checked all posts from each initiative published during June 2019. Our objective was to obtain a sample of the production from each of the fact-checking projects in order to understand their characteristics and particularities. There are six analysis categories: (1) discourse, (2) sources, (3) context, (4) classification, (5) graphic representation, and (6) funding. This analysis categories were established from a previous study (Storch et al. 2018), which mapped and analyzed similarities and differences in the verification methods adopted by Latin American fact-checking outlets.
The “discourse” category lists the discursive materials that fact-checking initiatives investigate. This analysis looks at the particularities of fact-checking in different social and political contexts, and in different media contexts.
The “sources” category references the sources of information that fact-checking initiatives use to perform their checks. The term “origin” refers to the original sources in the discourse that will be checked, this could be either a statement issued by an individual or organization or it could be the ones who circulate it. The sources of verification are what journalists use to verify data or information, and they then use that information as evidence to build their narratives on.
In order to build context we have to observe how fact-checking initiatives use verified information to reconstruct narratives. This involves looking at narrative strategies used to clarify misinformation from a particular discourse, taking into account what researchers describe as the “typology” of fake news (Tandoc et al. 2018a, 2018b; Wardle 2017).
Two other relevant issues (described in Section 3.2) are the ways in which information is classified and the graphic representations. Although each initiative has its own classification system, it is important to note that classifying information is one of the main reading references used for verifying circulated information, and must be able to provide adequate interpretations on the object being checked, so as not to be mistaken for misinformation. As Vizoso and Vásquez-Herrero (2019) state, graphic representations use elements such as color, icons, and texts to build up the meanings of verified information. Lastly, we analyze funding and the fundraising models used by the initiatives.
5 Data analysis
After examining the Portuguese language countries around the world we identified a total of 15 fact-checking initiatives, 13 of which are from Brazil and two from Portugal.[1] We did not found any active initiatives in Africa matching the geographical profile of CPLP countries. The overall data we collected shows that 86.6% of existing initiatives are linked to the professional fields of journalism or public media communications departments (13 initiatives). Nine of these are directly linked to media and journalism companies, one is a consortium of journalism vehicles, one is an individual project (Boatos.org), two are produced by public/governmental organizations. Only two initiatives have no direct link to journalism: one is a search engine developed by an online political marketing company (posverda.de) and the other is a website developed by a professional systems analyst (E-farsas). A detailed description of each category can be found in the research codebook.
Of particular interest is that 61.5% of all the journalism fact-checking initiatives are directly affiliated to consolidated news vehicles, and 46% work closely with conventional newsrooms. The only fact-checking consortium we found, the Projeto Comprova, is made up of consolidated news companies and a startup news company, Agência Lupa, which was being funded up until 2018 by Editora Alvinegra, the publishing agency which edits Piauí Magazine. Agência Lupa fact-checks for the Piauí’s Magazine’s website, as well as the UOL portal and Folha de S. Paulo newspaper, yet it does not share any editorial affiliations with any of these media companies (Table 1).
Fact-checking initiatives classification in Portuguese language countries.
| Initiative | Country | Affiliation | Classification | Organizational model linked to news media |
|---|---|---|---|---|
| Agência Lupa | Brazil | Piauí Magazine | Journalism company | News agency |
| Aos Fatos | Brazil | – | Journalism company | Fact-checking outlet |
| Boatos.org | Brazil | – | Individual project | Fact-checking outlet |
| E-farsas | Brazil | – | Individual project | Fact-checking outlet |
| Estadão Verifica | Brazil | Estado de S. Paulo | Journalism company spin-off | Editorial |
| Fato ou Fake | Brazil | G1 | Journalism company spin-off | Editorial |
| Fatos ou Boatos | Brazil | BNDES | Public organization | Public press office |
| Me engana que eu posto | Brazil | Veja Magazine | Journalism company spin-off | Editorial |
| O Observador | Portugal | O Observador | Journalism company spin-off | Editorial |
| Polígrafo | Portugal | – | Journalism company | Fact-checking outlet |
| posverda.de | Brazil | Cyberh company | Private organization | Search engine |
| Projeto Comprova | Brazil | 24 media vehicles | Journalism consortium | Consortium |
| Saúde sem Fake News | Brazil | Ministry of health | Public organization | Public press office |
| UOL confere | Brazil | Uol | Journalism company spin-off | Editorial |
| Verdade ou Boato | Brazil | GauchaZH | Journalism company spin-off | Editorial |
Another interesting detail (illustrated in Table 2) is the type of activity that each of these initiatives develop. Although they are all in the business of verifying information, they do differ in regards to the motives behind the fact-checks; for example, we found that some initiatives operate like newsrooms and are organized in a similar fashion to editorials or sections of an established media vehicle. 40% of the total initiatives we studied (6 initiatives) operate in this manner. We also found initiatives that operate like verification vehicles (4 initiatives, or 26.6%), whose entire editorial structure is focused on fact-checking. There is also a fact-checking consortium which adopts a cross-checking technique where the information is verified by journalists from different vehicles, and the verification result is made available to all consortium members.
Number of publications per initiative in June 2019.
| Initiative | Total N° Posts |
|---|---|
| Agência Lupa | 49 |
| Aos Fatos | 33 |
| Boatos.org | 74 |
| E-farsas | 48 |
| Estadão Verifica | 36 |
| Fato ou Fake | 30 |
| Fatos ou Boatos | 1 |
| Me engana que eu posto | 0 |
| O Observador | 11 |
| Polígrafo | 12 |
| posverda.de | 0 |
| Projeto Comprova | 0 |
| Saúde sem Fake News | 7 |
| UOL confere | 13 |
| Verdade ou Boato | 4 |
| Total | 318 |
Origin of discourses verified by Portuguese language initiatives.
| Initiative | Online content | Journalistic content | Public statements |
|---|---|---|---|
| Agência Lupa | 40 | 4 | 5 |
| Aos Fatos | 28 | 1 | 2 |
| Boatos.org | 74 | 0 | 0 |
| E-farsas | 48 | 0 | 0 |
| Estadão Verifica | 36 | 0 | 0 |
| Fato ou Fake | 30 | 0 | 0 |
| Fatos ou Boatos | 1 | 0 | 0 |
| O Observador | 9 | 0 | 1 |
| Polígrafo | 8 | 3 | 1 |
| Saúde sem Fake News | 6 | 0 | 0 |
| UOL confere | 11 | 0 | 2 |
| Verdade ou Boato | 4 | 0 | 0 |
| Total | 295 | 8 | 11 |
Agência Lupa is the only news agency that fact-checks for other news outlets, who then purchase the work from the agency. Two other initiatives are classified as “public press office” – they are based in the communication offices of two Brazilian public agencies (the National Bank for Social Development and the Ministry of Health) and fact-check the performance of the advisors. Lastly, we also found a search engine developed by a company specialized in digital political marketing, Cyberh, which users use to search information on the internet.
5.1 Discourses, sources and context
As we highlighted on the research methodology, this stage of the analysis uses Content Analysis (Bardin 2011; Bauer 2015). The information we analyzed was prepared from a systematic collection of all the fact-checking publications during the month of June 2019, a total of 318 published verifications.
5.1.1 Origin of the discourses
The first exploratory stage was to identify the standards involved in selecting the discourses which were to be verified by fact-checkers. From this data collection we were able to identify three main points of origin for the content verified by these initiatives: social networks and web pages, journalistic content and public statements (Table 3). The same information can circulate at the same time in more than one media environment – this happened on five occasions during our collection period, and as a result, they appear twice in Table 4. The overall majority of the fact-checks carried out by the analyzed agencies in June 2019 originated from social media platforms or web pages (295 posts, or 92.8%), mainly containing misinformation. This misinformation occurred in the form of text messages, images or videos which were widely circulated and reached a large number of people (Table 5).
Themes of publications verified by initiatives.
| Theme | Total N° posts |
|---|---|
| Politics | 182 |
| Entertainment and arts | 71 |
| Health | 19 |
| Environment | 13 |
| Legal | 11 |
| Economy | 9 |
| International | 8 |
| Safety | 5 |
Publications verified by initiatives according to typology of disinformation.
| Initiative | False content | Content with false context | Manipulated content | False profile |
|---|---|---|---|---|
| Agência Lupa | 16 | 12 | 9 | 2 |
| Aos Fatos | 19 | 5 | 4 | 0 |
| Boatos.org | 41 | 29 | 3 | 1 |
| E-farsas | 13 | 20 | 11 | 0 |
| Estadão Verifica | 24 | 8 | 4 | 0 |
| Fato ou Fake | 13 | 8 | 6 | 3 |
| Fatos ou Boatos | 1 | 0 | 0 | 0 |
| O Observador | 9 | 2 | 0 | 0 |
| Polígrafo | 1 | 2 | 1 | 0 |
| Saúde sem Fake News | 5 | 0 | 0 | 0 |
| UOL confere | 9 | 1 | 2 | 0 |
| Verdade ou Boato | 4 | 0 | 0 | 0 |
| Total | 155 | 86 | 40 | 6 |
Fact-check classifications in initiatives from Portugal.
| Polígrafo | Observadora |
|---|---|
| True | Right |
| True, but … | Almost right |
| – | Exaggerated |
| Inaccurate | – |
| – | Inconclusive |
| – | Misleading |
| False | Wrong |
| Pimenta na língua | – |
-
aWe refer to the Fact-check section of this media outlet.
Graphic representation used for classifying verified information.
| Initiative | Color | Icon | Text |
|---|---|---|---|
| Agência Lupa | ✓ | – | ✓ |
| Aos Fatos | – | ✓ | ✓ |
| Boatos.org | ✓ | – | ✓ |
| E-farsas | – | – | – |
| Estadão Verifica | – | – | – |
| Fato ou Fake | – | – | ✓ |
| Fatos ou Boatos | ✓ | ✓ | ✓ |
| Me engana que eu posto | – | – | – |
| O Observador | ✓ | ✓ | ✓ |
| Polígrafo | ✓ | ✓ | ✓ |
| posverda.de | – | – | – |
| Projeto comprova | ✓ | – | ✓ |
| Saúde sem Fake News | ✓ | – | ✓ |
| UOL confere | ✓ | – | ✓ |
| Verdade ou Boato | ✓ | – | ✓ |
Even still, journalistic content can also be identified as a source originating from the fact-checks, as is the case in 2.5% of the content in our analysis (8 posts). This is usually content from interviews given by public figures, but we also found journalist content which was contradicted by the fact-checks, although very little. Statements made by public figures on social networks or public statements by government representatives or members of the judiciary can also be verified – this occurred in 3.4% of the data in our analysis (11 posts). There was only one case (0.3% of posts), from the “Saúde sem Fake News” initiative, where we were unable to identify the origin of the verified information. In addition, the “Aos Fatos” initiative has published three fact-checks (0.9%) that originate from the initiative’s own verification processes, something which does not occur in any of the other projects. The four posts described here were excluded from the content analysis, which was carried out with a total of 314 posts.
There is also a high standardization of the themes in the materials selected for fact-checking. The prevalence of political content selected for verification (57.2% of the simple) is of note. The second most recurrent theme, representing 22.3% of the fact-checks, was called “varieties” which includes a wide range of gossip about celebrities and events which are classified in journalism as fait-divers. The definition of the publications themes was carried out considering the priority topic of verification. Each post was categorized by three independent coders and there were no significant differences in the delimitation of the themes (<5%). The cases in which there were discrepancies were reclassified considering the consensus among the coders.
Still in line with the origin of the discourses, we looked at the types of misinformation at the heart of the verified content. We looked at how the fact-checking initiatives defined each piece they verified and came up with four different types of misinformation: false content, content with a false context, manipulated content and false profile. False content is content that fact-checking initiatives can refute completely and correspond to 48.7% of the discourses. Examples include discourses made up by public authorities or images created about a situation that did not happen. Most of the content (151 out of 155 cases) is false information, but there are also false images (2 cases), false audio (1 case) and false video (1 case).
Content with false contexts is multimedia content that misinforms but really does exist, however, it is published in a false context either because it did not happen at the time stated or it did not happen in the way it is described. This type of misinformation represents 27% of the material in our analysis, most of which (75 out of 86 cases) are real videos that are published with out-of-context information. We found 29 cases of decontextualized images, 10 cases of decontextualized information, and one decontextualized audio. There is also manipulated content, which partially modifies real content for the purpose of changing its meaning. This type of misinformation appears in 12.6% of the content. The most common type of media to circulate manipulated content are images (26 out of 40 cases), followed by videos (11 cases), audio (2 cases) and textual information (1 case).
We find it important to point out that the analysis detected a pattern in relationship between the type of misinformation and the media to which it is most often linked: false information was circulated mainly in texts, while false contexts were mainly circulated in videos and images led to more manipulated content. There were six occasions where a profile was not verified by social media platforms and was found to be false. This happened in 1.9% of the publications we analyzed. There was only one case where we were unable to verify the type of disinformation. In addition, 27 publications were excluded from the analysis at this point because they were verified as “true” (9 cases) or verification was inconclusive (18 cases). The total of 287 publications were analyzed at this stage.
5.1.2 Checking sources and context
We also performed a second exploratory stage to identify patterns in the narrative construction of checks. We looked at the sources used to verify the content and the narrative structure that contextualizes the verification.
For the former, we mapped the sources used in the checks and presented to the reader. We consider checking the visible sources in the text to be relevant to fact-checking because these initiatives value the idea of “transparency” in editorial processes, and because these are the sources that journalists use as “elements of proof” when building their narratives for verification texts. There is a very wide variety of checking sources which can be classified according to their editorial positions in the checks: official sources, journalistic vehicles, and alternative sources.
The official sources are the characters/organizations in the fact-checked message who can offer reliable information (because they are the ones publishing the topic) about the statement and its context. We discovered three distinct manifestations from official sources: official documents (databases, streaming content, technical studies or reports, official or historical documents, law texts, administrative or judicial processes); communication advisors (informational or institutional content, notes and official statements, information sent directly to journalists, posts on social media profiles); and statements from individuals who hold official positions.
The outlets we analyzed regularly used journalistic vehicles as a source of verification. Previous reports, articles, interviews, and checks facilitate the contextual reconstruction which is needed to explain certain events and they help confirm the veracity of what is checked. We also identified the reference to the knowledge from experts: scientific entities and non-governmental organizations; international multilateral organizations; political parties; companies and entities not directly linked to the theme.
Our exploratory research has given us a systematic look at this element which can provide relevant data on the journalistic work performed by verification agencies. We have not delved into the details of this data in this article, but we consider this a relevant next step in the research. What really caught our attention was the volume of information where the only verified source came from journalists’ discussions with press offices – this was especially true for political issues. Self-reference is also an interesting fact – when building checks it is common for vehicles to use other checks previously carried out on similar subjects or materials published by the company that hosts the fact-checking initiative.
Lastly, another important element is the narrative format of the checks, or how they are presented to readers. This is what we label “context” in our analysis category. One of the fundamental distinctions between initiatives is whether they show (more or less explicitly) or do not show the original piece of disinformation. Some fact-checking initiatives are concerned that showing the piece of disinformation in the verification could lead to its dissemination. This is why “Estadão Verifica” never explicitly shows images or representations of the information that will be verified. All contextualization is done from the textual narration of the circulated content. Similarly, “Aos Fatos” also does not show the disinformation that will be verified, but all its material contains a link so readers can see prints of the newsroom content. Here readers are able to see how the misinformation is shaped, but they are not redirected to the agent who circulates it.
It is important to highlight that, even in cases where the disinformation piece is explicitly presented to the reader, it is always accompanied by some fact-check result definition (true, false, contradictory, and others). This can happen with text, like in “Polígrafo”, or with a visual mark (the verification labels are used to classify these kinds of cases), like in “E-farsas” and “Saúde sem Fake News”.
The initiatives that use visual marks (we explore this point below) for presenting fact-check results use narratives to present the content that will be checked (“Social networks are saying that … ”, “Someone said that … ”). Once these have been verified, they are given a label (“true” or “false”). After determining its veracity, the text will present the arguments from the original content and the verifications. If initiatives do not use any visual markings it is more often because the narratives in the text indicate the result of the verification (“It is false that … ”, “This is not … ”). This is followed up with the original context and the necessary explanations to prove what was verified.
5.2 Classification and graphic representation
The process of labeling fact-checked information depends on each initiative’s editorial structure and the type of discourse they verify. In some cases, as with Estadão Verifica, a section in the Estado de S. Paulo newspaper, classification labels are used together with an explanation of whether a fact-checked discourse is true or not. This is included in the body of the text and in the title. There are three initiatives that use this strategy in our analyzed corpus; two journalistic (“Estadão Verifica” and “Me Engana que eu Posto”) and one non-journalistic (“E-Farsas”). Most initiatives work with some form of classifying information.
This resource is especially relevant in fragmented circulation of fact-checks which occur on social networking sites when readers or the vehicle itself share content. But there is one initiative which classifies its information differently. The “Boato.org” initiative uses only one label to indicate whether the information conveyed is a rumor or not. It is an independent vehicle whose objective is to identify a list of hoaxes that occasionally circulate over the internet and then make that list available to readers.
Another common label, one which is used by four of the analyzed initiatives, is the true/false binomial, the objective of which is to label the discourse with an opposing classification – “one thing or the other”. The two press relations initiatives in our analyzed corpus use this type of classification. “Saúde sem Fake News”, linked to Ministry of Health in Brazil, classifies their fact-checks as “This information is true” and “This is fake news”, while “Fatos ou Boatos”, linked to the National Bank for Social Development (also a federal company) classifies their fact-checks as “Fact” or “Rumor”. There are two other Brazilian digital newspaper editorials which use the same binary classification: “Fato” or “Fake”, from the G1 newspaper, and “Falso” or “Enganoso”, from the UOL Confere initiative, linked to the UOL news portal. It is important to note that UOL Confere does not use this classification in all its fact-checks. One exception is when the site publishes fact-checks from Projeto Comprova (which it participates with). The criteria chosen to whether include the classification or not cannot be determined just by accessing the published data.
This type of classification helps to identify discourses based on factual elements, which are able to be verified objectively. However, it can “stifle” the fact-checking process, particularly because verified discourses tend to be more complex, and sometimes need further explanations in order to provide a satisfactory description of the situation. This is why the majority of fact-checking initiatives use broader classification structures.
We found initiatives in our analysis with classifications ranging from between three and 11 different labels. We could observe a standardization among all analyzed – all within the same general spectrum – with greater or lesser detail – ranging from “true/true” to “false/wrong. So, we identified intermediate levels of classification within each one of these labels which identify cases where parts of the discourse are correct, but not the entire discourse. The two Portuguese examples presented in the table below are relevant because they cover the general context of the classification models found by the research (Table 6). They have a similar classification system, creating a number of variants between right or true and wrong or false.
Some of the labels are comparable (in gray), but each initiative can establish different classification criteria. In the Polígrafo column a discourse can be deemed “inaccurate”, while in the Observador column that same discourse can be deemed “inconclusive” or contain “misleading” content. Additionally, the label “Pimenta na Língua”[2] in the Polígrafo column is a category created by this outlet, which describes it as “high level of falsehood. This classification is only attributed when the evaluated information is scandalously false or is a satire, published in a satirical space”.
Another interesting classification example is Projeto Comprova, a Brazilian fact-checking consortium. In addition to the most common labels, like true, false and misleading, this initiative also uses the labels “digitally altered”, “legitimate image” and “verified location”.
Lastly, another characteristic in the classification processes is the initiative’s use of exclusive labels for different types of discourse, such as statements made by public figures, verification of rumors, monitoring of public works, or political promises. The “Aos Fatos” initiative uses labels such as “inaccurate, exaggerated, unsustainable, and contradictory” are only applied to statements, while the “distorted” label is only applied to rumors.
The graphic representation is another relevant characteristic of the verification labeling process and, similar to classification indexes, they represent a quick understanding of the main value of the fact-check, particularly when it comes to true and false. According to Vizoso and Vásquez-Herrero (2019), the visual label is done through the combined use of elements such as color, icons and texts (Table 7). Most initiatives in our analysis (73.3%) used only text to build visual representations, while chromatic elements appear in 60% of the analyzed cases, and icons in only 26.63%.
It is not so common for the three elements above to be used at the same time – only three of the initiatives’ (20% of the total) graphics use color, icon and text (Fatos e Boatos, O Observador, and Polígrafo). Icons are often used to emphasize or reiterate the meanings that color and text have already conveyed.
Only Aos Fatos uses icons and text, but not color, to build its visual representations. It uses icons with words to help reinforce the meanings. Although black is also a color, its purpose is not to symbolize any specific meaning, unlike red and green in the images above. Conversely, black is used primarily to maintain a visual identity that “excludes” it from any categorizing function it may have. Another unique case is with the Fato ou Fake initiative. It only uses text as a graphic representation for classifications. The color gray is used to build a visual identification of the label to also exclude meaning by representing visual “neutrality”. Lastly, most of the initiatives analyzed (40%) use only color and text as visual resources for representing fact-checks.
5.3 Funding models
Looking at the funding models for each fact-checking initiative is relevant because it helps us understand not only the strategies behind the development of procedural and organizational innovations in journalism and the agents who develop them, but also the way in which these innovations are incorporated by consolidated media organizations. 40% of the cases we analyzed are fact-checking initiatives that are integrated with traditional journalism newsrooms in Brazil and Portugal: Me engana que eu posto (Veja Magazine), Verdade ou Boato (Gaúcha ZH newspaper), Estadão Verifica (O Estado de S. Paulo newspaper), Fato ou Fake (G1 news portal), O Observador and Uol Confere. These initiatives are funded by the vehicles that promote them, produced by journalists who work for the companies, and follow traditional funding models based on advertising and subscription sales. Many of these initiatives are also members of media conglomerates (such as the G1 portal, which is part of the Brazilian conglomerate Globo).
Communication advisors from Brazilian public agencies also seem to recognize the fact-checking format as a tool for producing information on the subjects that interest their advisers. The two Brazilian initiatives, Fato ou Boato (National Bank for Social Development) and Saúde sem Fake News (Ministry of Health), are funded by public money from the Federal Government’s budget. These initiatives are part of the public communication sectors that these entities belong to.
A significant percentage of fact-checking projects do not have any direct links to consolidated journalistic brands and presented financing models with different sources of revenue. Our analysis identified seven projects that fit this category (46.6%), as the table above illustrates. The types of funding described in the Table 8 were mapped empirically, by visiting the web pages of each of the initiatives, where the source of revenue for each was analyzed individually.
Summary of fact-checking initiatives that do not have any direct links to consolidated journalistic brands and presented financing models with different sources of revenue.
| Initiatives | Subscription | Donation | Sale of content | Training fee | Sale of services | Advertising | Investors |
|---|---|---|---|---|---|---|---|
| Agência Lupa | – | – | ✓ | ✓ | – | – | ✓ |
| Aos Fatos | ✓ | – | ✓ | – | ✓ | – | ✓ |
| Boatos.org | – | – | – | – | – | ✓ | – |
| E-farsas | – | – | – | – | – | ✓ | – |
| Polígrafo | – | – | – | – | – | – | ✓ |
| posverda.de | – | ✓ | – | – | ✓ | ✓ | ✓ |
| Comprova | – | – | – | – | – | – | ✓ |
Only one of the initiatives, Aos Fatos, uses reader subscriptions as a source of funding. The project’s overall content is free to access, but subscribers receive a weekly newsletter (called Aos Fatos +) which gives them behind the scenes information from the newsroom. In addition, subscribers can participate in monthly video debates with members of the production team. For Aos Fatos, subscriptions are a symbolic option – as of April 2020 the initiative had 37 subscribers, which represents a total revenue of US$ 172,17 per month (the amount Brazilian currency is R$ 990.00, converted into American dollars according to the exchange rate of March 30, 2021).
Another strategy used by fact-checking initiatives to ensure financial sustainability is the sale of content or services and charging a fee for training. Only three projects use this type of funding, all of them are Brazilian. A major portion of Aos Fatos’ revenues come from the sale of content (the fact-checks are published by other news vehicles or acquired by international news agencies) and the sale of specialized services. The Aos Fatos Lab is a branch of the agency that develops “sponsored artificial intelligence and automated fact-checking projects” and produces “thematic reports for the internal consumption of […] clients” (“Conheça o Aos Fatos Lab | Aos Fatos,” n.d., translated by authors).
Agência Lupa is structured like a news agency, and as such, the sale of content is a major part of its funding strategy. Even so, Lupa has developed a project called Lupa Educação, which “train citizens in fact-checking techniques”. This training occurs through workshops and lectures paid for by interested groups such as companies, universities, and journalism courses. Lupa Educação conducts training throughout Brazil, but some has also been conducted in Portugal and Spain.
Advertising, which is a standard form of financing for consolidated journalism companies, was used by only three initiatives in our analysis. Two of these initiatives (the E-farsas website and the Posverda.de search engine), which sell advertising spaces and use automated advertisements, are characterized in this paper as “non-journalistic” projects. The Boatos.org project is independently produced by journalists, its only source of funding coming from the sales of advertising space.
The main source of funding among the fact-checking initiatives analyzed in this paper is investors. Five of the seven projects using a hybrid financing model rely on the support of investors. The kinds of investors are interesting here since the projects do not predict large profits; they are experimental, have a relatively high operating cost, have difficulties guaranteeing revenues from the direct sale of their products, and there is a high level of competition. What’s more, they operate within a field (journalism) that is undergoing rapid transformations. Some of the main actors that support the fact-checking activities are Google and Facebook, who have an invested interest in developing innovative solutions in communication, especially to fight disinformation. Google and Facebook are partners with the Agência Lupa, Aos Fatos and Projeto Comprova initiatives and their international projects in support of journalism.
The following are individual investors: Agência Lupa’s first investor, Brazilian documentary filmmaker João Moreira Salles; the Portuguese company Polígrafo lists the Emerald Group as its shareholders, who are represented by its leader, N’Gunu Olívio Noronha Tiny; the B.creative media agency, founded by journalists Ricardo Fonseca and João Paulo Vieira; and journalist Fernando Esteves. It is interesting to note that the media mentions these individuals as a way to build their identity and credibility, in addition to building financial transparency. Moreira Salles is a widely renowned businessman, documentary filmmaker and cultural producer in Brazil and is viewed as a committed intellectual. In Portugal, N’Gunu Olívio Noronha Tiny is described as a “lawyer born in Angola, a banker and well-respected academic, a graduate of the Faculty of Law at the New University of Lisbon, a researcher for the London School of Economics and a Visiting Scholar at Harvard Law School”. Ricardo Fonseca and João Paulo Vieira are renowned journalists, known for their role in writing the magazine Visão (one of the few new magazines in Portugal). Lastly, Fernando Esteves is described as an “award-winning journalist, particularly for his work in journalistic investigations”, a writer and a university professor.
Another source of funding we identified in our analysis is donations, which only one of the initiatives currently uses: the posverda.de search engine. In addition to donations, posverda.de stated it sells services that specialize in “fake news advice”, it sells automated checking systems, and offers monetization services through cryptocurrency mining. The search engine was produced by the digital political marketing company Cyberh, based in the Brazilian capital of Brasília.
6 Discussion of data and conclusions
Fact-checking has been recognized by scholars, professionals, other media outlets as a tool for combating disinformation and misinformation, especially on social media platforms. Research on this issue seeks to understand the fundamentals and definitions of the practice in specific regions of the world, particularly the current situation in the United States and Europe. In a quick and superficial look at the wide range of journalistic fact-checking practices worldwide, we notice that the configuration of this kind of outlets is not entirely identical in different countries and continents, although there is a tendency to standardization.
In this article, we proposed an observation on the journalistic practice of fact-checking initiatives in Portuguese-speaking countries. The study offers a contribution to multilingual perspectives and tries to understand (RQ 1) what is the connection between fact-checking initiatives and the organizational models of the other media outlets in each Portuguese-speaking country. Also, the study addresses (RQ 2) the funding strategies of the fact-checking outlets and sought to understand (RQ 3) what discursive elements structure the verification methods of fact-checking outlets in Portuguese-speaking countries.
Among the data we analyzed, we have indicators of standardization in the organizational and production models of fact-checking outlets. The evidences that show standardization tendencies are related to the connection of these initiatives with traditional media. A large part of these initiatives (86.6%) are linked to professional practice in communication and journalism. Within this scenario, we noted that 61.5% of fact-checking initiatives have a direct relationship with consolidated news outlets, with 46% working in close collaboration with conventional newsrooms.
From the perspective of the material produced, we were able to identify the main origins of the content verified by these initiatives: social media and web pages, journalistic content, and public statements. During the analysis period, most of the fact-checking carried out by the analyzed outlets originated from social media platforms or web pages (295 posts, or 92.8%), containing mainly misinformation.
In summary, the contrasting aspects of the practices analyzed within the scope of this text are related to the classification models and the graphic representation created by the outlets. Most fact-checking initiatives use classification labels that vary among different degrees of correctness and incorrectness. Some also assume degrees of indeterminacy. We found outlets with classifications ranging from three to 11 different labels. We identified intermediate levels of classification within each of these labels that identify cases where parts of the discourse are correct, but not the entire discourse. In other words, the rating labels created by the outlets are composed of a specific set of variants between right or true and wrong or false.
Visually, classification labels are represented by icons, colors, and texts. A large portion of the initiatives use text (73.3%) to construct the visual representations of their classification labels. Chromatic elements appear in 60% of the cases analyzed, and icons in only 26.63%. Nevertheless, the elaborated representations are quite distinct from each other.
Gatekepping theoretical perspective was relevant for the analysis, as it allows us to consider the multiple influences on informational flows – and, therefore, on the institutionalization models of fact-checking. As Shoemaker and Vos (2009) argue, more than individual factors related to the journalist or editorial factors related to the characteristics of published information (organizational level), the gatekeeping perspective favors the recognition that aspects related to the social system and the institution also influence on the configuration of journalistic practices. It is in this sense that we understand the outstanding standardization processes recognized in fact-checking initiatives in Portuguese-speaking countries studied.
The present study corroborates research carried out in other geographic-cultural contexts (Kim and Buzelli 2022) that suggest the standardization of editorial processes related to fact-checking and a concentration of outlets of this type along with traditional models of journalism – either as editorials or as independent agencies linked to journalism structures already consolidated or derived from them. In this context, even though fact-checking initiatives can be understood as an editorial innovation, as they stimulate alternative revenue models and suggest relevant adjustments in the work dynamics of newsrooms, these characteristics of innovation are relatively demarcated by more complex institutionalization processes of fact-checking as part of journalistic culture.
Here we described aspects of the configuration of the methods and funding of fact-checking outlets and went a bit further in analyzing journalistic practices of verification. However, our analysis was partial. A systematic look at the production and circulation processes of these practices and also at the contexts in which they are embedded is still needed. Information control by governments or local political groups, competition in the news market, access to the internet and the digital literacy of the public, the costs of journalistic production, and organizational resistance to changes in the verification systems are all part of this context and demand attention from analysts and scholars as well.
In future studies, we intend to explore aspects of each step of the fact-checking process carried out by media outlets. Our aim is to develop a model of comparative analysis in which it is possible to understand and reveal the particularities of the practice and organization of fact-checking initiatives in each country around the globe.
The period chosen may affect the outcome of the analysis, since fact-checking initiatives, as well as journalistic coverage in general, is affected by the events on screen at the moment – electoral periods can be particularly significant in this context, since political content are relevant for verification by these agencies. Furthermore, it is relevant to consider the particularities of each member of the community of Portuguese-speaking countries. Distributed in different continents, with very divergent socio-political realities and with different media cultures. A more detailed analysis of the media contexts of each country can help us to explain the absence of fact-checking initiatives in most of them, as well as guide new research questions.
About the authors
Lucas Durr Missau PhD in Communication and Information from the Federal University of Rio Grande do Sul (UFRGS), Brazil. Adjunct professor at the Department of Communication Sciences and the Graduate Program in Communication, Federal University of Santa Maria (UFSM).
Laura Strelow Storch PhD in Communication from the National University of La Plata (UNLP), Buenos Aires, Argentina. Journalist and volunteer professor at the Department of Communication Sciences, Federal University of Santa Maria (UFSM).
References
Amazeen, Michelle. 2013. Making a difference? A critical assessment of fact-checking in 2012, vol. 0, (October), 1–40: New America Foundation. Available from: https://static.newamerica.org/attachments/10215-making-a-difference/Amazeen_-A_Critical_Assessment_of_Factchecking.4a3289cb3fcd4a8faf0d94f40ebcdc35.pdf.Search in Google Scholar
Amazeen, Michelle. 2015. Revisiting the epistemology of fact-checking. Critical Review 27(1). 1–22.10.1080/08913811.2014.993890Search in Google Scholar
Aos, Fatos. n.d. Conheça o Aos Fatos lab. https://www.aosfatos.org/aos-fatos-lab/ (accessed 24 June 2020).Search in Google Scholar
Bardin, Laurence. 2011. Análise de conteúdo, vol. 70. São Paulo: Edições.Search in Google Scholar
Barzilai-Nahon, Karine. 2011. Gatekeeping: A critical review. Annual Review of Information Science & Technology 43(1). 1–79. https://doi.org/10.1002/aris.2009.1440430117.Search in Google Scholar
Bauer, Martin. 2015. Análise de Conteúdo clássica: uma revisão. In Martin Bauer & George Gaskell (eds.), Pesquisa qualitativa com texto, imagem e som: um manual prático. Petrópolis: Vozes.Search in Google Scholar
Bode, Leticia & Emily Vraga. 2018. See something, say something: Correction of global Health misinformation on social media. Health Communication 33(9). 1131–1140. https://doi.org/10.1080/10410236.2017.1331312.Search in Google Scholar
Brandtzæg, Peter B. & Asbjørn Følstad. 2017. Trust and distrust in online services. Communications of the ACM 60(9). 65–71.10.1145/3122803Search in Google Scholar
Cappella, Joseph & Kathleen Jamieson. 1994. Broadcast adwatch effects a field experiment. Communication Research 21(3). 342–365.10.1177/009365094021003006Search in Google Scholar
Clayton, Katherine, Spencer Blair, Jonathan Busam, Samuel Forstner, John Glance, Guy Green, Anna Kawata, Akhila Kovvuri, Jonathan Martin, Evan Morgan, Morgan Sandhu, Rachel Sang, Rachel Scholz-Bright, Austin Welch, Andrew Wolff, Amanda Zhou & Brendan Nyhan. 2019. Real solutions for fake news? Measuring the effectiveness of general warnings and fact-check tags in reducing belief in false stories on social media. Political Behavior 42(December 2020). 1073–1095. https://doi.org/10.1007/s11109-019-09533-0.Search in Google Scholar
Dobbs, Michael. 2012. The rise of political fact-checking, vol. 0, (February 2012), 1–15. New America Foundation Available at: http://www.people-press.org/2013/08/08/amid-criticism-support-for-medias-watchdog-role-stands-out/%5Cnhttp://newamerica.net/sites/newamerica.net/files/policydocs/The_Rise_of_Political_Fact-checking.pdf.Search in Google Scholar
Fakida, Abdelrahman. 2021. Political fact-checking in the Middle East: What news can be verified in the Arab world? Open Information Science 5(1). 124–139. https://doi.org/10.1515/opis-2020-0117.Search in Google Scholar
García-Avilés, José A., Miguel Carvajal-Prieto, Alicia De-Lara-González & Félix Arias-Robles. 2018. Developing an index of media innovation in a national market: The case of spain. Journalism Studies 19(1). 25–42. https://doi.org/10.1080/1461670X.2016.1161496.Search in Google Scholar
Graves, Lucas. 2013. Deciding what’s true: Fact-checking Journalism and the new Ecology of news. New York: Columbia University.Search in Google Scholar
Graves, Lucas. 2017. Anatomy of a fact check: Objective practice and the contested epistemology of fact checking. Communication, Culture and Critique 10(3). 518–537. https://doi.org/10.1111/cccr.12163.Search in Google Scholar
Graves, Lucas. 2018. Boundaries not drawn: Mapping the institutional roots of the global fact-checking movement. Journalism Studies 19(5). 613–631. https://doi.org/10.1080/1461670X.2016.1196602.Search in Google Scholar
Graves, Lucas & Federica Cherubini. 2016. The rise of fact-checking sites in Europe. Oxford: Reuters Institute.Search in Google Scholar
Graves, Lucas, Brendan Nyhan & Jason Reifler. 2016. Understanding innovations in journalistic practice: A field experiment examining motivations for fact-checking. Journal of Communication 66(1). 102–138.10.1111/jcom.12198Search in Google Scholar
Kim, Bumsoo & Nicholas Buzelli. 2022. The logics of fact-checking website operations. Digital Journalism 1–24. https://doi.org/10.1080/21670811.2022.2089707.Search in Google Scholar
Lowrey, Wilson. 2017. The emergence and development of news fact-checking sites: Institutional logics and population ecology. Journalism Studies 18(3). 376–394.10.1080/1461670X.2015.1052537Search in Google Scholar
Nyhan, Brendan & Jason Reifler. 2010. When corrections fail: The persistence of political misperceptions. Political Behavior 32(2). 303–330.10.1007/s11109-010-9112-2Search in Google Scholar
Nyhan, Brendan & Jason Reifler. 2015. The effect of fact-checking on elites: A field experiment on U.S. state legislators. American Journal of Political Science 59(3). 628–640.10.1111/ajps.12162Search in Google Scholar
Poynter. n.d. Ethics policy. https://www.poynter.org/poynter-institute-code-ethics/(accessed 14 April 2020).Search in Google Scholar
Shoemaker, Pamela & Timothy Vos. 2009. Introduction: The significance of gatekeeping. In Gatekeeping theory, p. 9–16. London: Routledge.10.4324/9780203931653-5Search in Google Scholar
Spivak, Cary. 2010. The fact-checking explosion: In a bitter political landscape marked by rampant allegations of questionable credibility, more and more news outlets are launching truth-squad operations. American Journalism Review 32(4). 38+.Search in Google Scholar
Storch, Laura, Lucas Missau & Sabrina Cáceres. 2018. Mitômetro: A construção de um método de checagem em ambiente de aprendizado. Rebej 8(22). 69–86.Search in Google Scholar
Tandoc, Edson, Richard Ling, Oscar Westlund, Andrew Duffy, Debbie Goh & Lim Wei Zheng. 2018a. Audiences’ acts of authentication in the age of fake news: A conceptual framework. New Media & Society 20(8). 2745–2763. https://doi.org/10.1177/1461444817731756.Search in Google Scholar
Tandoc, Edson, Lim Wei Zheng & Richard Ling. 2018b. Defining “fake news”: A typology of scholarly definitions. Digital Journalism 6(2). 137–153. https://doi.org/10.1080/21670811.2017.1360143.Search in Google Scholar
Tandoc, Edson, Lim Wei Zheng & Richard Ling. 2020. Diffusion of disinformation: How social media users respond to fake news and why. Journalism 21(3). 381–398. https://doi.org/10.1177/1464884919868325.Search in Google Scholar
Uscinski, Joseph & Ryden Butler. 2013. The epistemology of fact checking. Critical Review 25(2). 162–180.https://doi.org/10.1080/08913811.2013.843872.Search in Google Scholar
Vizoso, Ángel & Jorge Vázquez-Herrero. 2019. Fact-checking platforms in Spanish. Features, organisation and method. Communication & Society 32(1). 127–142.Search in Google Scholar
Walter, Nathan, Jonathan Cohen, Lance Holbert & Yasmin Morag. 2019. Fact-checking: A meta-analysis of what works and for whom. Political Communication 0(0). 1–26. https://doi.org/10.1080/10584609.2019.1668894.Search in Google Scholar
Wardle, Claire. 2017. Fake news. It’s complicated. https://medium.com/1st-draft/fake-news-its-complicated-d0f773766c79 (accessed 27 May 2020).Search in Google Scholar
Wintersieck, Amanda. 2015. The rise and impact of fact-checking in U.S. Campaigns: Arizona State University.Search in Google Scholar
Wintersieck, Amanda. 2017. Debating the truth: The impact of fact-checking during electoral debates. American Politics Research 45(2). 304–331. https://doi.org/10.1177/1532673X16686555.Search in Google Scholar
York, Chance, James Ponder, Zach Humphries, Catherine Goodall, Michael Beam & Carrie Winters. 2019. Effects of fact-checking political misinformation on Perceptual accuracy and epistemic political efficacy. Journalism & Mass Communication Quarterly 1–23. https://doi.org/10.1177/1077699019890119.Search in Google Scholar
Young, Dannagal, Kathleen H. Jamieson, Shannon Poulsen & Abigail Goldring. 2018. Fact-checking effectiveness as a function of format and tone: Evaluating FactCheck.org and FlackCheck.org. Journalism & Mass Communication Quarterly 95(1). 49–75. https://doi.org/10.1177/1077699017710453.Search in Google Scholar
© 2022 the author(s), published by De Gruyter, Berlin/Boston
This work is licensed under the Creative Commons Attribution 4.0 International License.
