Bypassing Misinformation May Be As Effective As Correcting It

Misinformation is a persistent challenge for advocates working to shift public opinion and policy. The standard approach — debunking, or directly correcting false claims — is often effective, but it comes with real limitations. People generally don’t like being corrected, and some corrections can paradoxically strengthen memory for the very misinformation they’re meant to counter. Beliefs linked to conspiracy thinking are especially resistant to direct refutation since they’re often unfalsifiable. In this study, researchers explored whether a more indirect approach could be equally effective.

Their key insight draws on how beliefs and attitudes are psychologically organized. When attitudes rest on a single belief, directly correcting this belief would be the logical move. But attitudes can also be linked to multiple beliefs, and these beliefs are rarely active at the same time. This means that shifting which belief is top of mind may be enough to change people’s attitudes. The researchers tested a strategy they call “bypassing”: instead of directly confronting misinformation, this approach redirects attention to different beliefs whose implications run counter to the misinformation’s conclusion.

Across three experiments, they compared bypassing to a standard correction strategy and a control condition in which participants received only the misinformation. Experiments 1 and 2 recruited 360 and 303 participants, respectively, from an online research panel, while Experiment 3 used a representative U.S. sample of 772 participants.

All three studies used online questionnaires in which participants read simulated news articles about genetically modified (GM) crops. Each experiment followed the same basic design. Participants first read a false claim, then were randomly assigned to one of three groups:

• A group who read a bypassing article highlighting a different claim about GM crops;
• A group who read a correction article directly refuting the misinformation; or
• A control group who either received nothing further to read (Experiment 1) or read an unrelated article (Experiments 2 and 3).

Experiment 3 also added a fourth group of participants who only read unrelated articles and never received any misinformation.

In Experiments 1 and 2, the misinformation centered on GM corn products causing severe allergic reactions. Experiment 3 used a well-known but retracted study linking GM corn to tumor growth in mice. For Experiment 1, the bypassing message focused on GM crops’ potential to alleviate global hunger and malnutrition; for Experiments 2 and 3, it introduced the claim that GM crops could help save bee populations.

Across all three experiments, the researchers measured participants’ intentions to support policies restricting GM foods, their attitudes toward those policies, their belief in the misinformation, and their belief in the claim introduced by the bypassing article.

The results were consistent: compared to the misinformation-only control, both bypassing and direct correction significantly reduced support for GM food restrictions and produced less positive attitudes toward restrictive policies. In Experiment 3, results for both strategies were statistically indistinguishable from the no-misinformation control, indicating that both approaches fully neutralized the misinformation’s effect.

The two strategies appeared to work through different mechanisms. Direct correction reduced belief in the misinformation itself, while bypassing strengthened belief in the alternative claims, without necessarily reducing belief in the misinformation.

Notably, in Experiment 1, the bypassing belief (that GM foods can help feed the global population) was already widely held across groups. This suggests that simply reminding people of another belief they already hold can be enough to shift attitudes. Experiment 2 extended this finding by using a more novel bypassing belief (that GM crops can benefit bees), showing that the strategy also works when introducing a genuinely new belief rather than activating an existing one. Finally, by including a group who received no misinformation, Experiment 3 confirmed that the misinformation itself (that GM corn accelerates tumor growth in mice) was responsible for the elevated support for GM food restrictions and more favorable attitudes toward those restrictions.

All three studies measured only short-term outcomes, so it’s unclear whether these effects persist over time. The research was conducted exclusively in controlled online settings rather than in the field, and measured policy support intentions and attitudes, not actual behaviors. This limits how broadly the findings can be generalized. The authors also note that bypassing may be less effective when misinformation is highly accessible. In those cases, redirecting attention away from it may simply not work.

These findings offer a practical and underexplored approach for advocates navigating an environment full of entrenched misinformation. When it comes to issues like animal product consumption, the use of animals in entertainment, or keeping exotic animals as pets, direct debunking isn’t always the right tool, especially when false claims are embedded in someone’s identity or worldview. This research suggests that leading with compelling, evidence-backed claims that surface counter-beliefs may be just as effective as fact-checking, and potentially far less likely to trigger defensiveness.

In other words, rather than arguing against a false claim, advocates might make more headway by redirecting attention to a different belief that leads people toward the opposite conclusion intended by the misinformation. For audiences who hold conspiracy-adjacent beliefs or strong preexisting opposition, the bypassing approach may be especially well suited as it sidesteps the confrontation that can cause direct correction to backfire.

This summary was drafted by a large language model (LLM) and closely edited by our Research Library Manager for clarity and accuracy. As per our AI policy, Faunalytics only uses LLMs to summarize very long reports (~50+ pages) that are not appropriate to assign to volunteers, studies that contain graphic descriptions of animal cruelty or animal industries, and research on niche topics. We remain committed to bringing you reliable data, which is why any AI-generated work will always be reviewed by a human.