Are There Echo Chambers in the US News Ecosystem? Evidence From Twitter/X

TL;DR

This paper examines the presence of echo chambers in the US news ecosystem on Twitter by analyzing bias-labeled news accounts and their interactions, revealing distinct ideological clustering.

Contribution

It introduces a dataset of Twitter news accounts with bias labels and provides empirical evidence of echo chambers through network analysis.

Findings

Distinct clustering patterns indicating echo chambers

Limited interaction between conflicting ideologies

Bias influences information dissemination

Abstract

This study investigates echo chambers in social networks through an analysis of Twitter news accounts. Utilizing bias labels from the AllSides website, we construct a dataset representing six dimensions of news bias. Through manual extraction of follower/following relationships, we analyze interactions among 65 active Twitter news accounts. Despite the relatively small size of the network node data utilized, results reveal distinct clustering patterns indicative of echo chambers, with limited interaction between conflicting ideologies. This study underscores the potential impact of bias on information dissemination and democratic expression. These findings offer valuable insights into the dynamics of echo chambers in contemporary social media environments.