top of page

New Cx Research: Trust and Bias in News Evaluation

Recent research had shown that Americans believe that over half of news reporting is biased or inaccurate. Further, studies had shown that people’s evaluation of news articles as untrustworthy, biased, or inaccurate often depends on the political affiliation of the source and its alignment with the reader. This bias can result in differing evaluations from right- and left-leaning readers of the same article or claim.

Past research, however, did not distinguish between individuals’ trust in a news source, and their belief in the claims of the articles it publishes. Researchers also did not account for politically motivated responding, which may significantly affect such survey-based evaluations.

Cx PhD student Maurice Jakesch and faculty member Mor Naaman, along with collaborators from the Technion, addressed these issues in a randomized controlled experiment, using novel methods to separate the different factors that impact responses to news. The work will be presented at the 2019 Computation + Journalism Symposium and can be downloaded from SSRN.

The results of our study shows that the reputation and perceived politics of a media source is less important to evaluations of reporting than previously thought. In contrast to findings from previous studies, participants’ evaluations of headlines in this study were most influenced by the politics of the claim presented, and to a lesser degree by the politics and reputation of the media source. Our experimental results suggest that the salience of publisher demonstrated in past studies was possibly caused by conflation of article and publisher politics. These findings are consistent with the idea of motivated reasoning, where a reader’s people’s worldview affects their evaluation of information.

You can read more about the study in our Trust, Bias and News Evaluation Research Report, and in the Cornell Chronicle.


Recent Posts

See All

Cx PhD student Yiqing Hua has been working on using natural language processing techniques to detect propaganda in news. As a participant in the 2nd Workshop on NLP for Internet Freedom (NLP4IF), Y

bottom of page