Misinformation is more than "fake news": Using co-sharing to identify use of mainstream news for promoting misinformation narratives

Abstract

Most research concerning the volume and spread of misinformation on the internet measures the construct at the source level, identifying a set of specific "fake" news domains that account for a relatively small share of overall news consumption. This source-level categorization obscures the potential for factually true information from mainstream sources to be useful in the service of false or misleading narratives — a potentially far more prevalent form of misinformation. Using a combination of text- and network-analytic techniques, we find that articles from reliable sources that are co-shared with (i.e. shared by users who also shared) "fake" news on social media are significantly more likely to contain false or misleading narratives than articles from the same sources that are not co-shared. This is consistent with users strategically re-purposing information from mainstream sources to enhance the credibility and reach of misleading claims. Our frameworks broaden both the empirical and theoretical scope of misinformation research.