Effects of Misinformation on Online Discussions

CSH Friday talk

Jula Luehring

14 June 2024

Misinformation and emotions

Do (arousing) emotions make people believe in misinformation?

Emotional state

  • Replication study

  • False/accurate COVID-19 news headlines

  • Austria 2021

  • N = 422


\(\rightarrow\) No effects of emotional state on misinformation acceptance

Emotional response

\(\rightarrow\) More anger and less joy in response to false news

🗯 “Bullshit”, “Fake”

\(\rightarrow\) Function of emotion depends on existing beliefs

Problem #1


Different effects of emotions are overlooked by

  • mixing up different timings of emotions,

  • ignoring the function of emotions,

  • measuring positive and negative sentiment only.

Misinformation on social media

Collective dynamics online

  • Moralizing and arousing content gets high engagement

  • Misinformation: conflict and negative

    • But: only 0.3-6% in 5 studies from 2016-2021
    • Elite and ordinary partisan superspreaders

\(\rightarrow\) Misinformation is embedded in partisan intergroup dynamics

\(\rightarrow\) Secondary effects?

Nikolov et al., 2021

What are the effects of misinformation on discussions?

But: how to identify misinformation?

\(\rightarrow\) How do these choices influence downstream research results?

Problem #2


Misinformation is measured as clearly true or false instances,

  • neglecting less extreme types,

  • making it hard to isolate effects of misinformation.

Our objectives

  1. Collecting a systematic, large-scale and long-term data set for the German-speaking context

Continuous trustworthiness ratings by NewsGuard (#1)

  1. Approximating causal inference to test the effects of misinformation on emotions

Nonparametric matching strategy (#2)

Data collection

Posts from Twitter/X mentioning any of 347 German news domains


N = 9.3M discussions (20.6M tweets total)

93.8% trustworthy (>60)

Machine learning classification

pol_emo_mDeBERTa (Widmann & Wich, 2022; Macro F1=0.7)

Part I: Emotion in the post

Is trustworthiness associated with anger?


\(\rightarrow\) Trustworthiness predicts a 15% decrease in anger


But gray-area content matters, too!

Part II: Engagement

Is lower trustworthiness associated with higher engagement?

Models: Zero-inflated Negative Binomial (log-link)

Controls: PO, word count, following, initial emotions

\(\rightarrow\) 58% decrease in retweets and 43% decrease in quotes

Part III: Emotional responses

A) Correlations

Emotional response reflects emotion in post

\(\rightarrow\) Does trustworthiness actually affect emotional reactions?

B) Causal inference

Nonparametric matching


Nearest Neighbor and Mahalanobis distance

\(\rightarrow\) N = 87,132

Does trustworthiness affect emotional responses?




\(\rightarrow\) Less joy





\(\rightarrow\) 2% more anger

C) Direction of anger (TBD)

Out-group classification (Lasser at al., 2023; F1=0.8)

C) Origin of anger (TBD)

  • How do people talk in most angry discussions? Do they counterargue?
  • What are the topics in the first post?

  • Do the discussion networks differ (see Gonzalez-Bailon et al., 2010)?

Conclusion (tentative)

Emotions \(\rightarrow\) emotions \(\rightarrow\) engagement?

  • Sources with low trustworthiness predict anger
  • Emotions in discussions largely reflect emotions in initial post
  • Is engagement with low trustworthiness due to other factors?

\(\rightarrow\) Not the factfulness is harmful, but the content

\(\rightarrow\) Misinformation is a perfect tool to spread hateful content

Thank you!

Is lower trustworthiness associated with higher engagement?

Models: Zero-inflated Negative Binomial (log-link)

Controls: PO, word count, following, initial emotions