A group of faculty and student researchers at Clarkson University have published research aimed at assessing misinformation regarding COVID-19 across several different languages, nations, and social media platforms.
Golshan Madraki, assistant professor with the David D. Reh School of Business, said the group investigated COVID-19 misinformation and disinformation on social media in multiple languages and countries: Chinese (Mandarin) in China, English in the United States, and Farsi (Persian) in Iran. The study spanned multiple platforms, including Twitter, Facebook, Instagram, WhatsApp, Weibo, WeChat and TikTok.
Madraki said the research came about as she, Computer Science Professor Jeanna Matthews, and Assistant Professor of Electrical and Computer Engineering Yu Liu were collaborating on other projects.
“It was at the middle of the pandemic and all our meetings were through zoom with no actual social interactions. Therefore, during the first few minutes of our meetings, we were usually socializing and we always ended up talking about COVID and all the crazy nonsense that we have heard about COVID,” she said. “We noticed that since each of us were using different social media platforms, (i.e., Yu majorly uses Weibo and WeChat in Chinese; Jeanna majorly uses Twitter and Facebook in English; and, I majorly use WhatsApp, Telegram, Instagram, TikTok, and Facebook in Farsi), we were exposed to different types of misinformation that the other two have not even heard about it. This actually triggered us to search the literature for the differences of misinformation (particularly COVID related) on different social media platforms in different languages.”
The trio, accompanied by graduate student Jacqueline Otala and undergraduate student Isabella Grasso, defined misinformation, disinformation, mal-information and “infodemic” in the course of their research. Infodemic refers to the danger of misinformation during the pandemic.
The group explored why misinformation about the pandemic is so impactful, used sampling to log distinctive items of debunked misinformation, and used a qualitative approach to categorize the topics and roots of misinformation across the languages and platforms of focus.
According to Madraki, the group discovered that in all three countries, the source of the misinformation was the government or politicians. She also said English and Farsi misinformation samples have more in common in terms of the topic of misinformation than Chinese, specifically regarding the actions of the individual. Varying levels of government restriction on social media has an influence as well.
Madraki said while the Iranian government enforces some restrictions, China’s government influence is harsher, and Iran has more access to free internet, which sometimes exposes them to English misinformation.
The absence of misinformation with criminal roots and fewer categories of misinformation overall in Chinese social media is notable and points out the tradeoff in the control of misinformation, Madraki added. A key challenge moving forward, she said, will be in determining how to control misinformation without silencing the voices needed to hold governments accountable.
Madraki said sometimes it is difficult to detect misinformation from facts, particularly when the misinformation targets sensitive topics which could affect our judgment and sensibility. She also offered some advice on ensuring the information you are reading on social media is accurate.
“The next time you see content on your social media feeds that does not add up, please search about it,” Madraki said. “Fact-checking technologies have been advancing since the start of the pandemic. For example, Google search and its fact-checking technology can help a lot to recognize the misinformation and disinformation from facts.”
The full study, titled “Characterizing and Comparing COVID-19 Misinformation Across Languages, Countries and Platforms,” can be found at https://dl.acm.org/doi/pdf/10.1145/3442442.3452304.
