The Effects of Popularity Metrics in News Comments on the Formation of Public Opinion

Evidence from an Internet Portal Site

media user
popularity metrics
news comments
most-liked comments
bandwagon heuristics
online commenting behavior
data science
The Social Science Journal

Inyoung Park

Hyungbo Shim

Jang Hyun Kim

Changjun Lee

Daeho Lee*



The influence of online comment sections on the news has increased based on the development of collective online behaviors in the digitalized news media era. In this study, we focus on the effect of comment order (e.g., sorting comments by the number of likes or by the time of posting) on the formation of public opinion. We explore whether reading comments sorted by number of likes (a) induces more comments from users, (b) increases the expression of user opinions in response to others’ comments through the action of liking or disliking comments and (c) consolidates user opinion. For the empirical verification of the effects of popularity metrics, we chose a common topic (increasing minimum wage), collected actual data (reviewing 3,251 articles and the numbers of associated comments, likes, and dislikes), and compared news categories based on the existence of popularity metrics. Semantic network analysis was conducted with UCINET and python for K-means clustering, and cosine similarity. Our results show how the comment order in the internet news environment affects the commenting behavior of news consumers.