RESEARCH PAPERS

Can a 15-minute online game make people resilient against vaccine misinformation? Three preregistered experiments with 2,326 participants show that teaching manipulation technique recognition through gamified inoculation significantly improves detection of manipulative content — without making people more suspicious of reliable information.

Psychological Inoculation Improves Resilience Against Vaccine Misinformation

· 4 min read

Psychological Inoculation Improves Resilience Against Vaccine Misinformation

What if you could vaccinate people against misinformation the same way you vaccinate them against disease — by exposing them to a weakened dose?

That is the premise of inoculation theory, and a study published in Scientific Reports by Jon Roozenbeek, Sander van der Linden, and colleagues demonstrates that it works. Across three preregistered randomized controlled trials with 2,326 participants, a short online game called Bad Vaxx significantly improved people’s ability to spot manipulative vaccine content — without making them more suspicious of reliable information.

The Intervention: Bad Vaxx

The Bad Vaxx game is a gamified inoculation intervention that teaches players to recognize four manipulation techniques commonly used in vaccine misinformation:

  • Emotional storytelling — using anecdotes and emotional appeals instead of evidence
  • Fake expertise and pseudoscience — citing fabricated credentials or misrepresenting scientific findings
  • The naturalistic fallacy — claiming that “natural” is always better and vaccines are “unnatural”
  • Conspiracy theories — implying hidden agendas behind vaccination programmes

Rather than debunking individual myths (a game of whack-a-mole that never ends), the game teaches players to recognize the techniques behind the myths. This makes the intervention applicable to misinformation people have never seen before, as long as it uses one of the same manipulation strategies.

Two versions were tested: a “good” version where players fight misinformation, and an “evil” version where they play as misinformation spreaders. Both worked, but the “good” version consistently performed better.

Key Findings

Detection improves. Players rated manipulative vaccine content as significantly more manipulative than control group participants (Cohen’s d = 0.33 for the “good” version, 0.29 for “evil”).

No false alarm problem. The game did not make people more suspicious of reliable vaccine information. This is critical — an intervention that makes people distrust everything is worse than useless. Bad Vaxx selectively increases skepticism toward manipulative content.

Sharing quality improves. Players showed higher discernment in their sharing decisions — more willing to share non-manipulative content, relatively less willing to share manipulative content (Cohen’s d = 0.34 for the “good” version).

Technique recognition drives the effect. The game specifically improved participants’ ability to identify which manipulation technique was being used — not just whether something was true or false. This suggests people learned a transferable skill, not just facts about vaccines.

It works across the board. The effect did not vary significantly by political ideology or vaccination intentions, suggesting the game can reach a broad cross-section of the population.

Why Technique Recognition Matters

The finding about technique recognition is arguably the most important result. Misinformation evolves constantly — new myths replace old ones, narratives shift, and content adapts to current events. Debunking individual claims is necessary but insufficient; by the time a claim is debunked, three new ones have taken its place.

Teaching people to recognize manipulation techniques — emotional manipulation, fake expertise, naturalistic fallacies, conspiracy thinking — gives them a durable skill. It does not matter whether the misinformation is about vaccines, climate change, or elections. The techniques are the same.

This is exactly the approach that inoculation-based workshops at events like the Fake News Festival at Europa-Universität Viadrina aim to put into practice: letting participants experience these techniques in controlled settings so they can recognize them in the wild.

The Broader Inoculation Research Landscape

The Bad Vaxx game builds on a lineage of gamified inoculation interventions:

  • Bad News (2019) — the original game inoculating against general misinformation
  • Go Viral! (2021) — targeting COVID-19 misinformation
  • Cranky Uncle (2023) — specifically addressing climate change misinformation

Several of the researchers behind this work — particularly Jon Roozenbeek and Sander van der Linden from the University of Cambridge — are also co-authors of the recent papers on cyborg propaganda and malicious AI swarms. Together, these papers form a picture: the threat of disinformation is escalating through AI, but the psychological defenses — particularly inoculation — have solid empirical evidence behind them.

The Authors

  • Ruth E. Appel (Stanford University)
  • Jon Roozenbeek (King’s College London / University of Cambridge)
  • Rebecca Rayburn-Reeves (Duke University)
  • Melisa Basol (University of Cambridge)
  • Jonathan Corbin (Duke University)
  • Josh Compton (Dartmouth College)
  • Sander van der Linden (University of Cambridge)

Try It Yourself

The production version of the Bad Vaxx game is freely available at badvaxx.com. All data and analysis scripts are available on OSF.

References

  • Appel, R. E., Roozenbeek, J., Rayburn-Reeves, R., Basol, M., Corbin, J., Compton, J., & van der Linden, S. (2025). “Psychological inoculation improves resilience to and reduces willingness to share vaccine misinformation.” Scientific Reports, 15, 29830. DOI: 10.1038/s41598-025-09462-5