I was curious how the abstract goes about classifying fake news VS mainstream news (which they seem to imply means non-fake news)... > There exists no general

I was curious how the abstract goes about classifying fake news VS mainstream news (which they seem to imply means non-fake news)... > There exists no general consensus on a list of false news sources. We follow Allcott et al. (2019)’s compilation of websites (that they call “Fake News Sites”) that regularly produce false news stories > We further identify as mainstream news sources those that are acknowledged in surveys as authoritative websites that possess major national readership and produce false news content exceedingly rarely > To compile mainstream news sources, we searched under the keyword “news” in Alexa. From the results categorized under “news”, we selected for our study the top 12 websites in Alexa traffic I disagree with their classification method here, as well as their usage of the term fake news. Their classification model implies that all content emanating from fake news sources are false and that content emanating from mainstream sources are not false. For example in answering their primary questions (cited below) they assume that merely because someone has viewed a fake news website it implies that the content they consumed was false (or visa versa, that if the person viewed a mainstream website the content was true). Oh, and they dont define what true/false news content even means. > how prevalent is false news content? ... we study the behavioral antecedents behind news readership at the household level ... we prescribe how online platforms should tailor their content moderation policies to manage false news exposures at the individual household level So that last question really shows their hand imo. This entire abstract is about how the system can get people back onto the mainstream propaganda train - you know, where all the "true" news content can be found. Heavy eye roll.