With regards to the paper you posted, I don't think it does enough to escape the fallacy of correlation/causation. It stands to reason that a news event implica

With regards to the paper you posted, I don't think it does enough to escape the fallacy of correlation/causation. It stands to reason that a news event implicating refugees for a crime will attract anti-refugee sentiment, likewise why would such sentiment spread to unrelated news posts? Facebook has already been caught prioritizing controversial content in user news feeds to increase interaction. Facebook will quickly learn a user that displays such views interacts more with said content, and display more.