Abstract

A key factor in the dissemination of inauthentic information on social media is the interaction between human cognitive bias and algorithmic content curation. Research shows that some people have strong preferences for pro-attitudinal content. These tendencies are exacerbated by social media platforms, which often amplify low-quality content because it generates high levels of engagement among like-minded audiences. We propose using audience diversity as a quality signal to address this problem. To demonstrate the potential value of this approach, we combine a comprehensive dataset of news source reliability ratings compiled by domain experts with the web browsing histories from samples of U.S.-based Internet users. Our results indicate that partisan audience diversity is a potentially useful signal of higher journalistic standards. These results suggest that platforms should consider incorporating audience diversity into algorithmic ranking decisions.