Skip to content

Bad News: Building Resilience to Misinformation Through Inoculation

Speaker: Sander van der Linden
Affiliation: University of Cambridge, Social Decision Making Lab
Video ID: oJ2l48cNEJQ
Event: Academic lecture (venue unspecified)

TL;DR

Van der Linden presents psychological inoculation as an alternative to reactive fact-checking. Rather than debunking false claims after people have been exposed, inoculation preemptively builds resistance to misinformation techniques. His flagship intervention, the Bad News game, teaches six core manipulation tactics through interactive gameplay. The approach is demonstrated to work across demographics, scales to millions of users via platforms like YouTube and social media, and persists over weeks—addressing a critical gap in information integrity strategies where reactive fact-checking fails to prevent the "continued influence of misinformation."

Key claims

Fact-checking is ineffective because misinformation persists in memory. Once people integrate false information into their mental models, corrections don't fully erase it. The brain's memory network links the false claim to related knowledge, and even when a correction is acknowledged, people continue retrieving and reasoning from the original misinformation. This "continued influence effect" is why fact-checking alone has limited real-world impact.

Debunking amplifies the problem by repeating the myth. Fact-checks necessarily repeat the false claim when refuting it, increasing its familiarity and fluency. The brain mistakes fluency for truth—the more people hear something, the more likely they believe it. Therefore, standard debunking can inadvertently strengthen belief in the false claim it attempts to correct.

Most misinformation is subtle, not obviously false. The public fixates on outright hoaxes (flat earth, obviously fake videos), but the real problem is misleading content that contains a grain of truth but uses emotional manipulation or misleading framing. For example, a real news story about a doctor's death presented to imply causation with a vaccine is technically factual but deliberately misleading.

Inoculation theory provides a scalable alternative. Inspired by medical immunization, psychological inoculation exposes people to weakened doses of misinformation techniques in advance, triggering cognitive defenses. Two mechanisms operate: (1) an affective basis—forewarning people that they might be targeted activates psychological resistance, and (2) a cognitive basis—preemptively refuting weak examples of a technique builds mental "antibodies" that generalize to new attacks using the same tactic.

Six core manipulation techniques encode most misinformation: Polarization (pitting groups against each other), impersonation (fake experts, spoofed accounts, deepfakes), conspiracy thinking, trolling and ridicule, emotional exploitation (fear-mongering, outrage), and discrediting (attacking credibility of opponents or fact-checkers). If people are trained to spot these techniques, they develop broad-spectrum immunity to novel claims that employ them.

The Bad News game teaches inoculation through interactive simulation. Players adopt the role of disinformation spreader, learning to craft viral posts using the six techniques while managing credibility. The game presents fictional but realistic scenarios (based on actual events like the Warren Buffett Twitter impersonation, NASA parody accounts, and political conspiracy claims). Post-game, exposed players rate fake news as significantly less reliable, with effects that persist across weeks of follow-up testing.

Inoculation effects are robust across age, education, and ideology. Pre-post and randomized controlled trials show that both liberals and conservatives benefit equally, as do high and low education participants. This universal effectiveness is crucial for scaling interventions.

Long-term boosters maintain inoculation immunity. Initial effects decay without reinforcement, but repeated exposures sustain the benefit—paralleling medical vaccines. Regular testing or re-exposure to similar techniques maintains cognitive resistance over time.

Pre-bunking works at scale on social media platforms. Go Viral (a 5-minute game version for pandemic misinformation) and YouTube pre-bunking videos (30 seconds) demonstrate that shorter interventions can be deployed at scale. A large-scale YouTube ad experiment reaching ~1 million users showed modest but significant improvements in people's ability to identify manipulation techniques—a 5% boost, which exceeds typical advertising effectiveness and represents real-world validation of the inoculation framework.

Real-world application: Pre-bunking the Ukraine invasion. The U.S. and UK governments preemptively disclosed Russian invasion plans and predicted fake propaganda videos before the conflict began. By forewarning publics and NATO allies about the specific manipulation tactics they would encounter, they applied inoculation theory to geopolitical communication, potentially delaying the invasion and reducing belief in disinformation.

Methods and evidence

Inoculation theory origins: - Developed by social psychologist William McGuire in the 1960s, inspired by Cold War concerns about "brainwashing" of American POWs - Core idea: a weakened dose of persuasive appeal triggers refutational defenses - Applied to toothbrushing myths in lab experiments with early success

Bad News game design: - Social media simulator where players create posts, manage followers, and maintain credibility - Fictional but realistic scenarios illustrating each of the six techniques - Interactive choice: players generate their own misinformation, learning mechanics firsthand - Post-game quiz: players rate novel headlines for reliability

Empirical evidence: - Organic viral launch (15,000 participants): Within-subjects pre-post design. Fake news rated significantly less reliable after playing; real news unchanged. - Randomized trial (placebo control: Tetris game): Inoculation group found fake news less reliable and more confident in truth-discernment; control showed no change. - Longitudinal persistence: Effects maintained across 3+ weeks with repeated testing; decay observed when intermediate testing removed. - Cross-cultural validation (Sweden, Germany, Greece, Poland): Comparable effect sizes across countries. - YouTube intervention (1M+ exposure): 30-second pre-bunking videos on YouTube ads showed ~5% improvement in technique identification post-exposure.

Connections

Notes

This presentation complements and extends van der Linden's earlier 2026 video from the Kyiv School of Economics, offering more detailed walkthrough of the Bad News game mechanics, international applications, and large-scale social media deployments. The emphasis on practical game design and scalability is particularly valuable: inoculation theory only matters if it can reach millions, not just laboratory participants.

The distinction between active inoculation (where people create their own misinformation examples, as in Bad News) and passive inoculation (where corrections or counter-arguments are provided) is empirically important. Active versions consistently outperform passive ones, likely because generating one's own weakened dose requires deeper cognitive engagement and builds stronger ownership of defensive arguments.

The Ukraine case study represents one of the most significant real-world validations of inoculation theory outside the lab—a geopolitical operation explicitly designed using the framework. This elevates the work from academic curiosity to practical policy tool.

Finally, the YouTube results are noteworthy not for large effect sizes (5% is modest compared to lab effects), but for existence of effects at scale in a real-world distraction environment. The generalization from controlled settings to social media demonstrates that inoculation is robust enough for practical deployment.