Science Advances

paper

2020·Science(peer-reviewed)·science.org/doi/10.1126/sciadv.aay3539

Authors

Jennifer Allen·Baird Howland·Markus Mobius·David Rothschild·Duncan J. Watts

Credibility Rating

5/5

Gold(5)

Gold standard. Rigorous peer review, high editorial standards, and strong institutional reputation.

Rating inherited from publication venue: Science

Empirical study on misinformation prevalence in media diets, relevant to AI safety discussions about content moderation, algorithmic amplification, and understanding the actual scale of fake news problems that AI systems are designed to address.

Paper Details

Citations

413

40 influential

Year

2020

Methodology

peer-reviewed

Metadata

journal articleprimary source

Summary

This Science Advances study challenges the conventional narrative about fake news prevalence using a comprehensive, nationally representative dataset covering mobile, desktop, and television consumption. The research finds that fake news comprises only 0.15% of Americans' daily media diet, news consumption overall represents at most 14.2% of media consumption, and television accounts for roughly five times more news consumption than online sources. The authors argue that public misinformation and polarization are more likely attributable to the content of mainstream news or news avoidance rather than overt fake news, suggesting the scale of the fake news crisis has been significantly overstated.

Cached Content Preview

HTTP 200Fetched Apr 9, 20260 KB

# Evaluating the fake news problem at the scale of the information ecosystem
Authors: Jennifer Allen, Baird Howland, Markus Mobius, David Rothschild, Duncan J. Watts
Journal: Science Advances
Published: 2020-04-03
DOI: 10.1126/sciadv.aay3539
## Abstract

Mainstream news, mainly on television, vastly outweighs fake news, and news itself is a small fraction of U.S. media consumption.

Resource ID: c87e62324323ed83 | Stable ID: sid_PwWxK3QiHn