SCIENCE@home

Perverse Downstream Consequences of Debunking: Being Corrected by Another User for Posting False Political News Increases Subsequent Sharing of Low Quality, Partisan, and Toxic Content in a Twitter Field Experiment

M. Mosleh,Cameron Martel,Dean Eckles,David G. Rand

Published 2021 in International Conference on Human Factors in Computing Systems

ABSTRACT

A prominent approach to combating online misinformation is to debunk false content. Here we investigate downstream consequences of social corrections on users’ subsequent sharing of other content. Being corrected might make users more attentive to accuracy, thus improving their subsequent sharing. Alternatively, corrections might not improve subsequent sharing - or even backfire - by making users feel defensive, or by shifting their attention away from accuracy (e.g., towards various social factors). We identified N=2,000 users who shared false political news on Twitter, and replied to their false tweets with links to fact-checking websites. We find causal evidence that being corrected decreases the quality, and increases the partisan slant and language toxicity, of the users’ subsequent retweets (but has no significant effect on primary tweets). This suggests that being publicly corrected by another user shifts one's attention away from accuracy - presenting an important challenge for social correction approaches.

PUBLICATION RECORD

  • Publication year

    2021

  • Venue

    International Conference on Human Factors in Computing Systems

  • Publication date

    2021-05-06

  • Fields of study

    Computer Science, Political Science, Psychology

  • Identifiers
    DOI 10.1145/3411764.3445642
  • External record

    Open on Semantic Scholar

  • Source metadata

    Semantic Scholar

CITATION MAP

EXTRACTION MAP

CLAIMS

  • No claims are published for this paper.

CONCEPTS

  • No concepts are published for this paper.

REFERENCES

Showing 1-58 of 58 references · Page 1 of 1

CITED BY

Showing 1-83 of 83 citing papers · Page 1 of 1