The spread of fake news: a case study of the presidential elections of 2018 in Brazil
As fake news becomes more pervasive with the increasing adoption of digital platforms, understanding how disinformation spread and the factors that contribute to its continuation has become crucial, given the detrimental effects for democracies. This study investigates the spread of fake news during...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Estado: | Versión publicada |
| Fecha de publicación: | 2021 |
| País: | Brasil |
| Institución: | Fundação Getulio Vargas (FGV) |
| Repositorio: | Repositório Institucional do FGV (FGV Repositório Digital) |
| Idioma: | inglés |
| OAI Identifier: | oai:repositorio.fgv.br:10438/30800 |
| Acceso en línea: | https://hdl.handle.net/10438/30800 |
| Access Level: | acceso abierto |
| Palabra clave: | Fake news Disinformation Presidential elections Social media Sources Distribution platforms Spread Susceptibility Partisanship Desinformação Eleições presidenciais Mídias sociais Fontes Plataformas de distribuição Disseminação Suscetibilidade Partidarismo Administração de empresas Presidentes - Brasil - Eleições Eleições - Brasil - 2018 Redes sociais on-line |
| Sumario: | As fake news becomes more pervasive with the increasing adoption of digital platforms, understanding how disinformation spread and the factors that contribute to its continuation has become crucial, given the detrimental effects for democracies. This study investigates the spread of fake news during the Presidential Elections of 2018 in Brazil and how distinct social media and websites are used as distribution platforms and sources of disinformation. For such, a pre-existing data set of 346 fake news stories collected during the elections served as a starting point. Initially, through a reverse search process, the main websites responsible for disseminating disinformation were mapped. These sources were then analysed in terms of traffic and partisanship. Beyond a prevalence of right-wing fake news sources, a high concentration of web traffic was found. Five websites were responsible for almost 80% of all pageviews (or impressions) from all the 58 identified fake news sources. Furthermore, in order to investigate the circulation of disinformation on Facebook, Twitter and WhatsApp, the data set was filtered into the 58 most relevant unique fake news stories, which were later classified by political bias, engagement (number of shares), and segregated in four narratives. Firstly, it was found that all the analysed social media served as relevant distribution platforms for fake news, once 32 out of the 58 fake news stories circulated in all of them. Yet, Facebook was found to be more relevant than Twitter for that purpose. Secondly, the four major narratives that shaped the fake news stories were mostly related to an intense polarization and declining rates of trust in public institutions and media vehicles. Among these, fake news related to anti-left/anti-workers were predominant. Similarly to the first analysis, partisanship was noticeable during the spread of disinformation, as there were ten times more pro-Bolsonaro (or anti-Haddad) fake news stories than the polar opposite. Finally, the findings indicate that, while Facebook and Twitter were relevant distribution platforms, WhatsApp had a major impact on closed groups due to the reinforced cognitive effects and externalities that corroborate to the susceptibility and spread of fake news on social media. |
|---|