One in four of the most-viewed videos on coronavirus (Covid-19) in English on YouTube are factually incorrect, according to a new study that comes against the backdrop of several websites removing a 26-minute viral video called “Plandemic” that alleged that the disease was a conspiracy being promoted by people looking to profit from it.
Videos from government and professional sources on the disease scored high on facts but accounted for just 10% of the around 257 million views considered by the researchers from Canada whose study was published in the journal, BMJ Global Health on May 13.
“The general public has varying degrees of health literacy. When we consider the enormous volume and diversity of misleading information being generated and spread, it has significant potential for harm,” one of the study’s lead-author Heidi Oi-Yee Li, faculty of medicine, University of Ottawa, said over email.“Given that YouTube has grown since past outbreaks, its reach is greater than ever [the study says 2 billion users daily], and we felt it was important to evaluate whether misinformation was spreading to wider audiences.”
The researchers used keywords “coronavirus” and “Covid-19” separately to identify the top 75 most viewed that were in English, shorter than an hour in duration, and excluded non-audio and non-visual content.
They said 69 videos met the criteria, and together clocked 257,804,146 views. These were evaluated on facts, quality, reliability and usability. Videos made by the news media accounted for 47% videos.
“Our study highlights that over 25% of YouTube’s most viewed English videos contained non-factual or misleading information reaching over 62 million views,” read the study. “Professional and government videos demonstrated higher accuracy, usability and reliability across all measures but were largely under-represented in our sample.”
Subscriptions and presentation is what drove views, the author said.
“The most viewed video in the study with over 20 million views was produced by a YouTube and television celebrity. Meanwhile, the most popular government video only reached 4 million views,” wrote Heidi. “This is because governments typically use traditional information dissemination tools (e.g. infographics, posters, websites) which is less appealing and may not be as accessible for those who are less health-literate.”
The study is limited because of its focus on one language , but is one of the largest such exercises.
“It’s a fine study. Although ‘views’ don’t necessarily translate to ‘users’, and to calculate that you may need IP addresses that may not have been available, it does indicate the trend. The Indian context, however, will require a broader explanation,” said P Vigneswara Ilavarasan, professor at IIT-Delhi and social media researcher.
The problem may be far worse in India where such videos do the rounds as forwards on messaging platforms.
Indians rarely go to social media platforms directly to gain information, Ilavarasan said. “One reason is the legacy issue, where we tend to trust forwards from our elders and superiors without question. Then there is the lack of information literacy where people are unable to detect misinformation and are not good gatekeepers, and lastly there is manipulation where a video can be taken out of context,” said Ilavarasan.
“We’re committed to providing timely and helpful information at this critical time… We also have clear policies that prohibit videos promoting medically unsubstantiated methods to prevent the coronavirus in place of seeking medical treatment, and we quickly remove videos violating these policies when flagged to us,” said a YouTube spokesperson.