Emotional misinformation spreading

Apeksha Shetty, Jula Lühring, David Garcia, Annie Waldherr & Hannah Metzler
Complexity Science Hub Vienna & University of Vienna
Slides: hannahmetzler.eu/emomis_update_2024

2024-04-09

The Project

Introduction

• Emotions attract attention & increase sharing
• Beliefs are shaped by our social identity
• Group & emotion dynamics are central on social media
• Existing interventions focus on accuracy

Research questions

Observational social media analyses of emotional dynamics around news

Emotion regulation interventions

Agent-based modelling: network & algorithms

Results & Highlights (Part I)

Do fake news elicit emotions?

• COVID-19 news headlines
• Austria 2021
• Mood & response
  
• Response:
  • More anger
  • Less joy

Do emotions increase susceptibility?

• Higher anger in people good & bad at recognizing false news
• Angry responses because most people recognize false news
• Emotions arise when information contradicts existing beliefs

Theory: Change of problem understanding

• Emotions are functional
• Humans are not gullible: motivated belief vs. manipulation
• Misinformation as symptom of real-world societal problems
• Motivated partisan minority shares misinformation
• Majority: uninformed, little news interest & silent

\(\rightarrow\) Questioning the role of emotions

\(\rightarrow\) Our solution: Targeting social dynamics & open-mindedness

• reduced trust in politics, science, institutions
• loss of privilege/status

Results & Highlights (Part II)

Measuring misinformation: NewsGuard

Expert ratings of news sources on 9 journalistic quality criteria

• Trustworthiness score per source
• In-depth analysis
• Increasing inclusion of low-trustworthy news
• Includes German sources since 2019

Emotions around news on Twitter

Discussion threads below tweets linking to news sources

• German speaking
• Emotions:
  Machine Learning

Causality: Matching, dataset with trustworthy and non-trustworthy articles that are very similar

Emotions depend on news trustworthiness

Low-trustworthy news: more anger & less joy (preliminary!)

Social identity intervention

• Goal: Increasing skepticism
• How: Beliefs incongruent with social identity
• Complementary & Alternative Medicine (CAM) vs. right-wing extremism
• Intervention shows link with
  • historical information
  • news articles

Intervention effect on trust & sympathy

Highlighting inconsistency of beliefs with social identity reduces trust & sympathy in alternative medicine.

Algorithm effects on news trustworthiness

Agent-based model of US politicians’ retweeting behavior

• most of the news are trustworthy
• Most effective: down-ranking low trustworthy posts
• Downranking anger: increases trustworthy posts
• Combining interventions is best

Other highlights

• 3 supervised internships + 2 Master theses
• Presentations at leading conferences: ICA Toronto 2023, MISDOOM Amsterdam 2023, SPSP 2022 (online), …
• EMOMIS Video for the Digital Humanism series
• Chapter in a public report on climate misinformation
• Media mentions (APA, DerStandard, Kurier, Science.ORF, FM4 news, Woman Balance…)

Upcoming outreach highlights

• Upcoming talks
  • klimaaktiv (Bundesministerium für Klimaschutz, May)
  • Volkshochschule (September)
• Upcoming media highlights
  • ORF documentary
  • DERSTANDARD Podcast
  • Article in Spektrum der Wissenschaften

Thank you!

• Project website: hannahmetzler.eu/emomis
• Preprint: Emotions in misinformation studies.
• Paper: Social drivers & algorithmic mechanisms on digital media

Appendix

COVID-misinformation items

• Actual true & false COVID-19 headlines
• From fact-checking websites & mainstream news sources
• Accuracy ratings