New Study: User-based surveillance is not effective against fake-news
The defence against fake news, to a large extent, relies on the social media users identifying and "flagging" suspicious stories, but are we, the users, really capable of doing that? No, a new study concludes.
Fake news and disinformation are challenging societies, democracies and economies. While the number of people who prefer social media (SOME) as their primary news source seems to increase, an efficient defence against fake or misleading stories has still not appeared.
Social media platforms rely on users “flagging” inappropriate online content, but according to a study, conducted by assistant professor Michele Coscia and associate professor Luca Rossi from The IT University of Copenhagen, the "flagging-concept" does not offer an efficient identification of fake news. In the research article Distortions of political bias in crowdsourced misinformation flagging, recently published in the English Journal of The Royal Society Interface, the two researchers analyse, which stories get flagged as fake.
- The results are disturbing. We show how expecting users to flag content carries the problematic assumption that users will genuinely attempt to estimate the veracity of a news item to the best of their capacity. Even if that was a reasonable expectation to have, users’ estimation of veracity will be made within their individual view of the world and variable polarization, Michele Coscia says.
Most of us tend to have a slightly distorted view of the things and point of views, that harmonises with our perception of the world and this, combined with the polarized dynamic of content circulation within social networks, could, according to Michele Coscia, create a context where actual fake news are being left unverified as long as they do not leave their friendly part of the network.
- Fake-news require careful fact-checking, and this takes time. Given the amount of information that is constantly shared online, the problem is mainly how we can ensure to fact-check the stories that we have reasons to believe are fake? Designing screening systems, such as those based on users flagging, that does not consider what we know about users’ behavior and social dynamics, is just doomed to fail. If we want to prevent fake news from spreading, we need to understand the social dynamics supporting the process, Michele Coscia says.
Learn more about the study Distortions of political bias in crowdsourced misinformation
Luca Rossi, Associate Professor, phone +45 7218 5036, email firstname.lastname@example.org
Jari Kickbusch, phone 7218 5304, email email@example.com