AustraliaA study conducted by Meta and academic researchers analyzed the impact of removing 'untrustworthy' content from Facebook and Instagram during the 2020 U.S. election. The study found that exposure to content from repeated falsehood sources was rare, averaging 1.1% on Facebook and 0.1% on Instagram. However, this content was highly concentrated, with 23% of Facebook users receiving 80% of such content. When posts from these sources were removed from the feeds of over 8,000 users, there was no measurable change in their beliefs or attitudes. The study highlights challenges in measuring social media's role in shaping political views, as users often select their own information sources. Researchers note that while the results seem reassuring, the methodology has limitations, particularly regarding how 'untrustworthy' sources are defined.
Bias read (Center): The article presents a balanced overview of the study's findings, acknowledging both the apparent reassurance that misinformation exposure is rare and the deeper concerns about measurement accuracy and definition of 'untrustworthy' sources. While the study is funded by Meta, the analysis does not sl
