The study smacks of elitism, gotcha questions and tailored outcomes. You are comparing regular NY Times, NPR, Boston Globe readers, which is an extremely biased

The study smacks of elitism, gotcha questions and tailored outcomes. You are comparing regular NY Times, NPR, Boston Globe readers, which is an extremely biased sample - above average education, affluent and engaged to people who read news on SM casually. Look at two of the questions from the "facts" based survey: https://www.journalism.org/2020/01/24/election-news-pathways-project-frequently-asked-questions/#measuring-overall-political-knowledge https://www.journalism.org/2020/01/24/election-news-pathways... (1) - Whether the unemployment rate in the U.S. has gone up, down or stayed the same since Donald Trump took office (gone down) (2) - What determines the number of votes a state has in the Electoral College (the number of seats the state has in the U.S. House and Senate) These are extremely technical questions. In (1) are they talking about the unemployment rate till any point on time, which is a random variable or the average? A gotcha questions In (2), I do read an hour of two of news everyday (I know, it is a problem) and did really well on American History, but that question would still take me a few moments to ponder. The causality is backwards - people who read SM to get news are less engaged and the people who read the establishment media are more engaged. Like the cigarette industry publishing an ad on the efficacy of cigarettes as an health aid.