CoChat

Evaluating the fake news problem at the scale of the information ecosystem

Open access
Allen, Jennifer, Howland, Baird, Mobius, Markus, Rothschild, David, Watts, Duncan J. · 2020 · Science Advances

The DOI matches its Crossref record; 21 of the 21 references it cites were found in a registry or an index.

Abstract

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

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1126/sciadv.aay3539

Cites

21 of 21 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Overall information consumption by category and platform over time, from January 2016 to December 2018. Breakdown of consumption for (A) the entire adult sample, 18 years and older, (B) 18 to 24 years old, and (C) 55 years and older. See table S6 for numerical values.
Fig. 1 Overall information consumption by category and platform over time, from January 2016 to December 2018. Breakdown of consumption for (A) the entire adult sample, 18 years and older, (B) 18 to 24 years old, and (C) 55 years and older. See table S6 for numerical values.CC BY-NC · PubMed Central
Fig. 2 Detailed breakdown of overall media consumption for Online and TV. (A) Online consumption (including mobile and desktop) for the top 2000 sites per applications on Comscore. (B) TV consumption by program category. Total online consumption is 227 min per person per day, of which 58% is accounted for by the top 2000 sites, while total television consumption is 232 min per person per day. To compute news consumption in search and social media, excluding YouTube, we use the share of referrals from the site in question that redirect to news articles as a proxy for the share of time a user is exposed to news-related content on the platform. For YouTube, which does not redirect users to external sites, we randomly sampled 10,000 videos per month (weighted by viewing time) and computed the percentages that were classified as “news and politics”. Because portals such as MSN, Yahoo, and AOL almost always display some news-related stories on their landing pages, we count 100% of time spent on portals as news consumption. Last, news consumption in the “variety” category of television viewing is computed as 100% of time attributed to late-night comedy programs, such as The Daily Show With Trevor Noah, which are known to contain commentary on politics and current events. For clarity, (A) shows only the top 15 of 28 categories (see table S7 for numerical values).
Fig. 2 Detailed breakdown of overall media consumption for Online and TV. (A) Online consumption (including mobile and desktop) for the top 2000 sites per applications on Comscore. (B) TV consumption by program category. Total online consumption is 227 min per person per day, of which 58% is accounted for by the top 2000 sites, while total television consumption is 232 min per person per day. To compute news consumption in search and social media, excluding YouTube, we use the share of referrals from the site in question that redirect to news articles as a proxy for the share of time a user is exposed to news-related content on the platform. For YouTube, which does not redirect users to external sites, we randomly sampled 10,000 videos per month (weighted by viewing time) and computed the percentages that were classified as “news and politics”. Because portals such as MSN, Yahoo, and AOL almost always display some news-related stories on their landing pages, we count 100% of time spent on portals as news consumption. Last, news consumption in the “variety” category of television viewing is computed as 100% of time attributed to late-night comedy programs, such as The Daily Show With Trevor Noah, which are known to contain commentary on politics and current events. For clarity, (A) shows only the top 15 of 28 categories (see table S7 for numerical values).CC BY-NC · PubMed Central
Fig. 3 Television versus desktop news consumption aggregated over all age categories 18 to 55+. For each month, the overlap panelists are separated into groups corresponding to different ranges of web news consumption. For each group, the mean television news consumption and group size as a percentage of all panelists are computed. Overtime averages for the mean television news consumption and size of each group are calculated by computing the mean television news mean and mean group size over all 36 months. Error bars are SEs obtained via bootstrapping for group size and group television news consumption, respectively, and are smaller than the symbols. See fig. S2 for all results broken down by age group, and tables S8 and S9 for numerical values.
Fig. 3 Television versus desktop news consumption aggregated over all age categories 18 to 55+. For each month, the overlap panelists are separated into groups corresponding to different ranges of web news consumption. For each group, the mean television news consumption and group size as a percentage of all panelists are computed. Overtime averages for the mean television news consumption and size of each group are calculated by computing the mean television news mean and mean group size over all 36 months. Error bars are SEs obtained via bootstrapping for group size and group television news consumption, respectively, and are smaller than the symbols. See fig. S2 for all results broken down by age group, and tables S8 and S9 for numerical values.CC BY-NC · PubMed Central
Fig. 4 News-only consumption by age. Detailed breakdown of news-only consumption by age group for (A) online (including mobile and desktop) and (B) television. See fig. S4 (A and B) for results plotted over time from January 2016 to December 2018. See table S11 for numerical values.
Fig. 4 News-only consumption by age. Detailed breakdown of news-only consumption by age group for (A) online (including mobile and desktop) and (B) television. See fig. S4 (A and B) for results plotted over time from January 2016 to December 2018. See table S11 for numerical values.CC BY-NC · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 21 references

Research works printed but not linked: 2

  1. D. J. Watts D. Rothschild Don’t blame the election on fake news. Blame it on the media. Columbia Journalism Review (2017); https://www.cjr.org/analysis/fake-news-media-election-trump.php.
  2. S. Athey M. M. Mobius J. Pál The Impact of Aggregators on Internet News Consumption (2017); https://papers.ssrn.com/abstract=2897960.

Cited by

Metadata

Authors
Allen, J., Howland, B., Mobius, M., Rothschild, D., & Watts, D. J.
Affiliations
  • Massachusetts Institute of Technology (US)
  • Harmony Labs, 311 W 43rd St., New York, NY 10036, USA.
  • Microsoft Research New England, 1 Memorial Dr., Cambridge, MA 02142, USA.
  • Microsoft Research New York City (United States) (US)
  • University of Pennsylvania (US)
Corresponding author
  • Allen, Jennifer
  • Howland, Baird
  • Watts, Duncan J.
Venue
Science Advances
Published
2020-04-03
Volume
6
Issue
14
Pages
eaay3539-eaay3539
ISSN
2375-2548 (ISSN-L)
OpenAlex
W3014767596
Cited by
528
References
21
Citations by year
2026: 52 · 2025: 65 · 2024: 101 · 2023: 108 · 2022: 99 · 2021: 74 · 2020: 26
FWCI
102.92
Type
journal article
Language
en
Open access
Licence
cc-by-nc
Locations
  • Science Advances (publishedVersion, cc-by-nc)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by-nc)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • ScholarlyCommons (University of Pennsylvania) (submittedVersion, cc-by-nc)
  • Scholarly Commons (University of Pennsylvania) (submittedVersion, cc-by-nc)
Funders
The Nathan Cummings Foundation (18-08129)
Topics
  • Social Sciences › Sociology and Political Science › Misinformation and Its Impacts
  • Social Sciences › Sociology and Political Science › Media Influence and Politics
  • Social Sciences › Communication › Communication and COVID-19 Impact
Keywords
fake news, misinformation, news consumption, media diet, television news, online news, media avoidance, political polarization, democracy, information ecosystem
MeSH terms
Adolescent, Adult, Communications Media, Evaluation Studies as Topic, Female, Humans, Male, Mass Media, Middle Aged, Surveys and Questionnaires, Television, United States, Young Adult, Social Media

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.