Investigating Misinformation Amid a Pandemic
A spotlight on Mozilla Fellow in Residence Oleg Zhilin
Like everyone else, Oleg Zhilin had different plans for 2020.
As a Mozilla Fellow and Masters of Computer Science student at Montreal’s McGill University, Zhilin planned to continue his research into misinformation, online platforms, and Canadian elections. After a salvo of misinformation upended the U.S. election in 2016, Zhilin and his colleagues at McGill had been examining how the phenomenon was playing out in their home country.
“We wanted to analyze and understand the online dynamics surrounding our election,” Zhilin explains. The findings could then help Canadian lawmakers prepare protections for future elections.
In 2019, Zhilin and his colleagues made headway on their research. The verdict: “Election misinformation exists in Canada, but not as significantly as in the U.S.,” Zhilin says. (Money in politics is a big reason: Canada’s leading political party spends about as much on political advertising as a single U.S. state.)
Then, in early 2020, COVID changed the research agenda. “We essentially dropped everything we were doing and focused our attention on the pandemic,” Zhilin recalls.
Over the course of the next several months, Zhilin and his fellow researchers aimed their expertise and energy at understanding how misinformation and COVID were intersecting. They combed through countless public posts, hashtags, and news articles related to the virus.
One portion of their research, published in June, explored how online conversations about the pandemic were shaping Canadians’ beliefs and behaviors. The essay was titled “The causes and consequences of COVID-19 misperceptions: Understanding the role of news and social media,” and published in the Harvard Kennedy School Misinformation Review.
“Misinformation can affect the transmission pattern of a pandemic, which is concerning for social media platforms that are increasingly being relied upon for news consumption,” Zhilin explains. “It is difficult to estimate the level of individuals' exposure to false or misleading claims because platforms try to detect and remove such content before it captures the attention of their users.”
Misinformation can affect the transmission pattern of a pandemic.
Zhilin adds: “Even if we accept that an individual has consumed misinformation through social media or traditional news, we still need to know whether they believe it and will change their behavior in response. This essay was one step by our research team towards the clarification of these dynamics with respect to COVID-19 in a Canadian context.”
Another project by Zhilin and his colleagues investigated government communication surrounding the pandemic. “We were trying to figure out: what is the best way to reach an audience that has a certain level of disbelief?”
More recently, Zhilin and his team investigated the hesitancy toward vaccines that some Canadians post online. “In anticipation of this becoming a bigger conversation in the coming months, we want to be able to make recommendations,” he explains. “Like: What are typical Canadians' concerns regarding vaccines? And, How can you best address them?”
Throughout the course of research, Zhilin and his colleagues took an open-source approach. They participated in the Digital Ecosystem Research Challenge, which encourages computer scientists, political scientists, and other students across Canadian institutions to pool their expertise. “The idea was to distribute data to teams across Canada,” Zhilin explains.
Over the course of the year, Zhilin and his colleagues uncovered common threads linking different research pursuits. Canadians consume information on social media differently than they do traditional media, Zhilin explains — people have visceral reactions to posts and often share them having read only the headline.
It’s a phenomenon they intend to study further, especially when it comes to vaccines. “Our ongoing work aims to clarify what types of vaccine-related content captures Canadians' attention. So far, we have some indication that a news article referencing controversial stances will generate more conversation than an article on vaccine development progress.”