The prevalence of misinformation on social media


\(\rightarrow\) how harmful is misinformation on social media then?

Figure 1. Exposure to elite misinformation is associated with sharing lower-quality outlets and with conservative estimated ideology (Mosleh & Rand, 2022).

1=false information without intention to deceive (Wardle & Derakhshan, 2017);
2depending on measurement and topic (Scharfbillig et al., 2026)

Effects of misinformation beyond belief


Individual-level experiments show

But they suffer from [low ecological validity]

\(\rightarrow\) study effects of misinformation where it spreads

1=based on circumplex model (Russell, 1980) and
constructivist view (Leach & Bou Zeineddine, 2021; Lindquist & Barrett, 2012)

Effects of misinformation beyond exposure


Observational social media data shows misinformation

But it suffers from [low internal validity]

  • confounded by other source cues, hateful and sensationalist language, …

\(\rightarrow\) calls to rethink causality and estimate effects in the wild (Budak et al., 2024; Lorenz-Spreen et al., 2022; Tay et al., 2024)

What are the effects of misinformation on (collective) online interactions?

Data collection


  • Systematic and large-scale data set for the German-speaking context: 9.9M news tweets, 11M replies (Oct 2020 – Mar 2022)

Figure 2. Data collection and pre-processing pipeline.

Dataset descriptives


Figure 3. Proportions of untrustworthy (<60; purple) and trustworthy (>60; green) news tweets.



  • robust 6% of untrustworthy news

Figure 4. Probability of each news tweet containing each emotion (0-1), predicted by pol_emo_mDeBERTa2 (Widmann & Wich, 2022).




  • anger and fear are overall higher across the whole observation period

Figure 5. Correlations between the variables in news posts and discussions.
  • emotional spillover into the discussions

Data collection (2)


Figure 6. Data collection and pre-processing pipeline.

Matching as a middle ground


  • keeps the naturalistic context of platform data

  • builds a synthetic control group comparable on observed confounders \(\rightarrow\) blocks backdoor paths (Ho et al., 2007; Pearl & Mackenzie, 2019)

DAG bias Political bias T Untrustworthy news bias->T Y Engagement & reply emotions bias->Y author Author (followers, following, tweets) author->T author->Y post Post (topic,word count, emotions) post->T post->Y T->Y

Figure 7. Directed Acylic Graph (DAG) with matched; treatment; outcome.

Matching as a middle ground


  • keeps the naturalistic context of platform data

  • builds a synthetic control group comparable on observed confounders \(\rightarrow\) blocks backdoor paths (Ho et al., 2007; Pearl & Mackenzie, 2019)

  • evaluate performance based on differences


Figure 8. Matching with Nearest Neighbor and Mahalanobis distance (Ho et al., 2025).

Confounder-adjusted effects on engagement


Figure 9. Marginal effects estimated by zero-inflated negative binomial models (bootstrapped 95% CIs).


\(\uparrow\) more retweets for tweets with untrustworthy news sources

\(\downarrow\) but fewer likes & replies

Confounder-adjusted effects on emotions


Figure 10. ATE estimated by linear regression (95% CI represent 10,000 bootstraps).


\(\uparrow\) more anger, disgust and fear

\(\downarrow\) and less joy in response to untrustworthy sources

Self-selection dynamics

Figure 11. Panel a: CDF of reply contributions (dashed = 80% threshold); Panel b: Anger for top contributors (accounting for 80% of replies).

Self-selection dynamics

Figure 11. Panel a: CDF of reply contributions (dashed = 80% threshold); Panel b: Anger for top contributors (accounting for 80% of replies).

\(\rightarrow\) angrier users preferentially engage, creating self-reinforcement

DAG bias Political bias T Untrustworthy news bias->T Y Engagement & reply emotions bias->Y author Author (followers, following, tweets) author->T author->Y post Post (topic,word count, emotions) post->T post->Y topic Unobserved topic->T topic->Y M Self-selection, curation T->M M->Y

Figure 12. Directed Acylic Graph (DAG) with matched; dashed = unobserved; who replies = part of the effect, not matched.

Conclusion

Limitations:
- Source \(\neq\) claim: we estimate effects of source trustworthiness, not of false claims in single posts as well as excluding non-textual content (Nenno et al., 2026);
- Unobserved confounders: topic, timing, algorithmic exposure \(\rightarrow\) tweet-level topics, causal sensitivity analyses, debiasing;
- Missing data: deleted tweets and suspended accounts

Thank you!



Email: jula.luehring@gesis.org

Bluesky: @julaluehring.bsky.social

Website: julaluehring.github.io

References

Allen, J., Howland, B., Mobius, M., Rothschild, D., & Watts, D. J. (2020). Evaluating the fake news problem at the scale of the information ecosystem. Science Advances, 6(14), eaay3539. https://doi.org/10.1126/sciadv.aay3539
Bak-Coleman, J., Lewandowsky, S., Lorenz-Spreen, P., Narayanan, A., Orben, A., & Oswald, L. (2025). Moving towards informative and actionable social media research (arXiv:2505.09254). arXiv. https://doi.org/10.48550/arXiv.2505.09254
Baribi-Bartov, S., Swire-Thompson, B., & Grinberg, N. (2024). Supersharers of fake news on Twitter. Science, 384(6699), 979–982. https://doi.org/10.1126/science.adl4435
Budak, C., Nyhan, B., Rothschild, D. M., Thorson, E., & Watts, D. J. (2024). Misunderstanding the harms of online misinformation. Nature, 630(8015), 45–53. https://doi.org/10.1038/s41586-024-07417-w
Ecker, U. K. H., Lewandowsky, S., Cook, J., Schmid, P., Fazio, L. K., Brashier, N., Kendeou, P., Vraga, E. K., & Amazeen, M. A. (2022). The psychological drivers of misinformation belief and its resistance to correction. Nature Reviews Psychology, 1(1), 13–29. https://doi.org/10.1038/s44159-021-00006-y
Goldenberg, A., Garcia, D., Halperin, E., & Gross, J. J. (2020). Collective Emotions. Current Directions in Psychological Science, 29(2), 154–160. https://doi.org/10.1177/0963721420901574
González-Bailón, S., Lazer, D., Barberá, P., Godel, W., Allcott, H., Brown, T., Crespo-Tenorio, A., Freelon, D., Gentzkow, M., Guess, A. M., Iyengar, S., Kim, Y. M., Malhotra, N., Moehler, D., Nyhan, B., Pan, J., Rivera, C. V., Settle, J., Thorson, E., … Tucker, J. A. (2024). The Diffusion and Reach of (Mis)Information on Facebook During the U.S. 2020 Election. Sociological Science, 11, 1124–1146. https://doi.org/10.15195/v11.a41
Guess, A. M., Malhotra, N., Pan, J., Barberá, P., Allcott, H., Brown, T., Crespo-Tenorio, A., Dimmery, D., Freelon, D., Gentzkow, M., González-Bailón, S., Kennedy, E., Kim, Y. M., Lazer, D., Moehler, D., Nyhan, B., Rivera, C. V., Settle, J., Thomas, D. R., … Tucker, J. A. (2023). How do social media feed algorithms affect attitudes and behavior in an election campaign? Science, 381(6656), 398–404. https://doi.org/10.1126/science.abp9364
Ho, D., Imai, K., King, G., & Stuart, E. A. (2007). Matching as Nonparametric Preprocessing for Reducing Model Dependence in Parametric Causal Inference. Political Analysis, 15(3), 199–236. https://doi.org/10.1093/pan/mpl013
Ho, D., Imai, K., King, G., Stuart, E., Whitworth, A., & Greifer, N. (2025). MatchIt: Nonparametric Preprocessing for Parametric Causal Inference.
Juul, J. L., & Ugander, J. (2021). Comparing information diffusion mechanisms by matching on cascade size. Proceedings of the National Academy of Sciences, 118(46), e2100786118. https://doi.org/10.1073/pnas.2100786118
Lazer, D., Hargittai, E., Freelon, D., Gonzalez-Bailon, S., Munger, K., Ognyanova, K., & Radford, J. (2021). Meaningful measures of human society in the twenty-first century. Nature, 595(7866), 189–196. https://doi.org/10.1038/s41586-021-03660-7
Leach, C. W., & Bou Zeineddine, F. (2021). A Systems View of Emotion in Socio-political Context. Affective Science, 2(4), 353–362. https://doi.org/10.1007/s42761-021-00051-z
Lindquist, K. A., & Barrett, L. F. (2012). A functional architecture of the human brain: Emerging insights from the science of emotion. Trends in Cognitive Sciences, 16(11), 533–540. https://doi.org/10.1016/j.tics.2012.09.005
Lorenz-Spreen, P., Oswald, L., Lewandowsky, S., & Hertwig, R. (2022). A systematic review of worldwide causal and correlational evidence on digital media and democracy. Nature Human Behaviour, 1–28. https://doi.org/10.1038/s41562-022-01460-1
Lühring, J., Shetty, A., Koschmieder, C., Garcia, D., Waldherr, A., & Metzler, H. (2024). Emotions in misinformation studies: Distinguishing affective state from emotional response and misinformation recognition from acceptance. Cognitive Research: Principles and Implications, 9(1), 82. https://doi.org/10.1186/s41235-024-00607-0
McLoughlin, K. L., Brady, W. J., Goolsbee, A., Kaiser, B., Klonick, K., & Crockett, M. J. (2024). Misinformation exploits outrage to spread online. Science, 386(6725), 991–996. https://doi.org/10.1126/science.adl2829
Mosleh, M., Cole, R., & Rand, D. G. (2024). Misinformation and harmful language are interconnected, rather than distinct, challenges. PNAS Nexus, 3(3), pgae111. https://doi.org/10.1093/pnasnexus/pgae111
Mosleh, M., & Rand, D. G. (2022). Measuring exposure to misinformation from political elites on Twitter. Nature Communications, 13(1), 7144. https://doi.org/10.1038/s41467-022-34769-6
Nenno, S., Puschmann, C., Fuławka, K., & Lorenz-Spreen, P. (2026). Political disinformation concentrates in repeated claims across media (p6yh9_v3). SocArXiv.
Pearl, J., & Mackenzie, D. (2019). The Book of Why. Penguin Books.
Petersen, M. B., Osmundsen, M., & Arceneaux, K. (2023). The “Need for Chaos” and Motivations to Share Hostile Political Rumors. American Political Science Review, 1–20. https://doi.org/10.1017/S0003055422001447
Robertson, R. E., Green, J., Ruck, D. J., Ognyanova, K., Wilson, C., & Lazer, D. (2023). Users choose to engage with more partisan news than they are exposed to on Google Search. Nature, 618(7964), 342–348. https://doi.org/10.1038/s41586-023-06078-5
Russell, J. A. (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39(6), 1161–1178. https://doi.org/10.1037/h0077714
Scharfbillig, M., Lewandowsky, S., Altay, S., Van, A. M., Kozyreva, A., Hertwig, R., Lorenz-Spreen, P., Diresta, R., Valenzuela, S., Egidy, S., Quattrociocchi, W., & Orben, A. (2026). Fractured reality. https://doi.org/10.2760/9358883
Tay, L. Q., Hurlstone, M., Jiang, Y., Platow, M. J., Kurz, T., & Ecker, U. K. H. (2024). Causal inference in misinformation and conspiracy research. Advances.in/Psychology, 2, e69941. https://doi.org/10.56296/aip00023
Vaccari, C., & Chadwick, A. (2020). Deepfakes and Disinformation: Exploring the Impact of Synthetic Political Video on Deception, Uncertainty, and Trust in News. Social Media + Society, 6(1), 2056305120903408. https://doi.org/10.1177/2056305120903408
Vosoughi, S., Roy, D., & Aral, S. (2018). The spread of true and false news online. Science, 359(6380), 1146–1151. https://doi.org/10.1126/science.aap9559
Wardle, C., & Derakhshan, H. (2017). Information Disorder: Toward an Interdisciplinary Framework for Research and Policymaking. Council of Europe Report. https://doi.org/http://tverezo.info/wp-content/uploads/2017/11/PREMS-162317-GBR-2018-Report-desinformation-A4-BAT.pdf
Widmann, T., & Wich, M. (2022). Creating and Comparing Dictionary, Word Embedding, and Transformer-Based Models to Measure Discrete Emotions in German Political Text. Political Analysis, 1–16. https://doi.org/10.1017/pan.2022.15