The «broadcast» model no longer works in an era of disinformation

Country/area: United Kingdom

Organisation: First Draft

Organisation size: Small

Publication date: 18 Dec 2020

Credit: Carlotta Dotto, Rory Smiths, Chris Looft

Project description:

2020 has shown us that the top-down broadcast model of the communications ecosystem is broken. It is networked and we are failing to understand how existing communities online have been hijacked to become vectors for mis- and disinformation. Using the ‘great reset’ conspiracy theory as a case study, we looked at the extent to which online communities have been used to become spreaders of disinformation following a decentralized approach, providing evidence on why disinformation travels quicker.

Impact reached:

The piece had a meaningful impact on the debate around new approaches to fight online disinformation and the role of fact-checks. It was shared by high-profile journalists, disinformation experts and researchers in the field, and its findings were cited in related articles, including by the leading Brazilian news website TAB from UOL, with 100 million users per month and 7.5 billion page views per month. The reaction to the piece was overwhelmingly positive and stimulated conversations around the role of pre-bunks and debunks in fighting disinformation. 

In 2020 disinformation has become more pervasive than ever. The piece proposes an innovative approach to a crucial issue that has a large impact on our daily life. Using data to find evidence of the power imbalance between conspiracy theories and the fact checks, our piece highlighted how often disinformation and conspiracy theories like the Great Reset have substantial structural advantages in the way they travel compared to quality information: instead of following the top-down or linear model of the mainstream media and fact-checks, they are shared across networked, participatory, and alternative information ecosystems.


Techniques/technologies used:

We used CrowdTangle’s API to collect 11,808 posts from Facebook groups and unverified Facebook pages that mentioned “great reset” or “greatreset” between November 16, 2020 and December 6, 2020. We then filtered for posts that included outward links to other websites and social media platforms.

To understand the potential impact of fact-checks on the conspiracy theory we also identified all the URLs from reliable news organizations that debunked the conspiracy theory and ran those URLs through CrowdTangle’s links endpoint to gather all the Facebook posts that shared them.

These two datasets, which included a total of 7,775 posts, were then merged for analysis using Pandas and Gephi. In Gephi we re-sized the nodes based on the in-degree values in order to highlight the URLs that were most frequently shared in this set. URLs with more shares are represented by larger nodes. We used the «ForceAtlas 2» layout to bring accounts that frequently share among each other closer together. 


What was the hardest part of this project?

The most challenging part of the project was the development of a new methodology to accurately analyse how pieces of information travel across different communities online. We decided to look at networked data to illustrate the dynamics that reinforce conspiracy theories such as the Great Reset, with an eye on the impact of fact checks and debunks published by trusted sources. The use of network analysis tools like Gephi can be time-consuming too.



What can others learn from this project?

Understanding how disinformation travels is crucial to develop tactics to deal with its spread. Our research proved that we need to move beyond thinking of disinformation as something broadcasted by bad actors down to unquestioning readers: false and misleading information circulates and evolves within robust ecosystems, where uncertain users collectively participate in a new sense-making. This suggests that instead of “chasing” misinformation and individual pieces of misleading narratives, we should instead focus to pre-empt its emergence, creating initiatives and participatory strategies with existing communities.

In an another point highlighted in the article, we mention the “data deficits,” situations of crisis in which there is a high demand for information about a topic but the supply of credible information is low. These deficits are quickly filled up by misinformation and conspiracy theories as people try to make sense of the situation. Identifying and preventing the emergence of this lack of information could be an important tool for journalists to prevent the spread of disinformation.

Project links: