Improving Readers' Awareness of Divergent Viewpoints by Displaying Agendas of Comments in Online News Discussions

Abstract

Readers of online news articles often hold related discussions in social media, sharing comments with various news-relevant topics, which we denote as 'agendas'. Online discussions that span diverse agendas can help readers mitigate the biased influences from news media. However, displaying comments based on popular voting often fails to reveal such diversities to readers. In this paper, we designed a proof-of-concept Hagendas that aims to improve awareness of divergent agendas in online news discussions. It presents possible agendas automatically derived from news articles, and enables readers to view and filter the comments accordingly. We evaluated how Hagendas would affect users' online news discussion practices through a within-subjects experiment with 95 online participants from MTurk. While the agenda tags and filtering features did not significantly increase the number of distinct agendas people identified through reading comments, 77.9% of participants preferred having those features for exploring online news discussion.