In a new study by the Oxford Internet Institute, University of Oxford analysis shows that Covid related misinformation videos primarily spread through social media, with Facebook the primary channel for sharing misinformation without sufficient fact checks in place to moderate content.
The Oxford study, ‘Covid-related Misinformation on YouTube’, examined over a million Covid-related YouTube videos which circulated on social media and identified videos YouTube had eventually removed because they contained false information. It finds that Covid-related misinformation videos do not find their audience through YouTube itself, but largely by being shared on Facebook. Data analysed by the Oxford researchers shows YouTube misinformation videos were shared nearly 20 million times on Facebook during between October 2019 and June 2020. They had a higher reach on social media than the five largest English-language news sources on YouTube, CNN, ABC News, BBC, Fox News and Al Jazeera combined, whose videos were shared 15 million times.
Facebook also generated a higher number of reactions than other social media platforms. Misinformation videos shared on Facebook generated a total of around 11,000 reactions (likes, comments or shares), before being deleted by YouTube. In comparison, videos posted on Twitter were retweeted on average around 63 times.
The Oxford researchers also found that out of the 8,105 misinformation videos shared on Facebook between October 2019 and June 2020, only 55 videos had warning labels attached to them by third party fact checkers, less than 1% of all misinformation videos. This failure of fact-checking helped Covid-related misinformation videos spread on Facebook and find a large audience.
Oxford researchers observed that despite YouTube’s investment in containing the spread of misinformation, it still took YouTube on average 41 days to remove Covid-related videos with false information. Misinformation videos were viewed on average 150,000 times, before they were deleted by YouTube.
Dr Aleksi Knuutila, Postdoctoral Researcher, Oxford Internet Institute, said:
“People searching for Covid-related information on YouTube will see credible sources, because the company has improved its algorithm. The problem, however, is that misinformation videos will spread by going viral on other platforms, above all Facebook.
“The study shows that misinformation videos posted on YouTube found a massive audience, and it is likely to have changed people’s attitudes and behaviour for the worse. For the sake of public health, platform companies need to ensure the information people receive is accurate and trustworthy.”
Other findings include:
Oxford researchers analysed 1,091,8756 distinct Covid-related misinformation videos posted on Facebook, Twitter and Reddit during the period October 2019 and June 2020. The full report can be found here.
Former Research Associate
Aleksi Knuutila is an anthropologist and data scientist. He was working as a Postdoctoral Researcher in the Computational Propaganda programme, where he used computational and qualitative methods to study political communications and communities.
Research Associate
Aliaksandr is a political communication researcher and was Director of Research at the Programme on Democracy and Technology at the Oxford Internet Institute.
DPhil Student
Hubert is a DPhil student at the OII, interested in computational propaganda and social movements online.
Research Associate
In 2022 Jonathan Bright became the Head of AI for Public Services at the Turing Institute, having previously been a faculty member of the OII. A political scientist, he specialises in computational and ‘big data’ approaches to the social sciences.
Professor of Internet Studies
Philip N. Howard is a professor of sociology, information, and international affairs. He is Director of the Programme on Democracy and Technology, and is a Professorial Fellow of Balliol College.
This project will focus on how bots, algorithms and other forms of automation are used by political actors in countries around the world.
