Curiosity changes minds where facts and arguments fail

1

The scandal that held

A fabricated 2014 canvassing study was honestly rebuilt and still produced measurable, durable attitude change from one structured conversation.

McRaney opens with graduate student Michael LaCour’s claim, published in Science, that a single twenty-minute talk with a gay canvasser could shift support for same-sex marriage. Two researchers, David Broockman and Joshua Kalla, tried to extend the work and found the survey data had been fabricated; the paper was retracted.

Instead of abandoning the question, they ran their own honest experiment with real canvassers in Miami. Attitudes toward transgender rights shifted after a single ten-minute conversation when canvassers used a precise sequence: ask the person to recall a time they were judged unfairly, then link that memory to the group under discussion.

The episode establishes the book’s method. McRaney follows the replication trail rather than repeating viral headlines. Dramatic clean results in social science deserve suspicion until independently checked; the truth about belief change is procedural and messier than any single study suggested.

Facts alone did not move the Miami voters. A structured emotional exercise did. That distinction becomes the thread the rest of the book pulls.

In practice: Broockman and Kalla’s honest Miami canvass produced lasting transgender-rights shifts after ten-minute conversations built around a personal memory of unfair judgment.

2

Corrections rarely backfire

Backfire is real only under narrow identity-threat conditions; most people, most of the time, update at least slightly toward accuracy when given clear corrections.

McRaney revisits the 2010 Nyhan and Reifler study that made the backfire effect famous: conservative subjects shown corrections about Iraq’s weapons of mass destruction grew more convinced the weapons existed. The result spread as proof that facts are useless.

Later, larger replications—including work by Thomas Wood and Ethan Porter across thousands of subjects and dozens of political claims—found backfire is the exception, not the rule. Most people do move a little toward accuracy.

What the original study likely captured was narrower: corrections that threaten identity, delivered without any acknowledgment of the person’s underlying concern, can trigger defensive reasoning. Facts work weakly and unreliably when the belief functions as a badge of group membership rather than a claim about the world.

The obstacle is not that human reasoning is broken. People are often answering a different question—will accepting this cost me standing with people I trust?—and answering it correctly.

In practice: Wood and Porter’s large replications across dozens of claims showed ordinary corrections usually produce small updates toward accuracy rather than entrenched denial.

3

Beliefs as membership badges

Many contested beliefs are held to signal group membership, so changing them threatens relationships and belonging, not just opinions.

McRaney argues that a large share of fights over vaccines, climate, or political figures are affiliative rather than epistemic. He uses Dan Kahan’s identity-protective cognition framework to explain why more numerate people can be worse at reasoning about charged facts: they have more tools to defend a group-consistent conclusion.

Once a belief fuses with self, family, and community, an attack on the claim registers physiologically like an attack on the person. Research on social rejection activating the same brain regions as physical pain helps explain why people will sacrifice accuracy to preserve belonging.

Persuasion is therefore a social problem before it is an informational one. You cannot out-argue someone’s need for community. Effective approaches must offer a path to revise the belief without forcing an exit from the group, or must face the emotional cost of that exit directly.

Contempt for the belief is correctly read as a threat to the believer’s whole social world, which guarantees resistance. Protecting social bonds is, in this framing, a reasonable priority rather than simple irrationality.

In practice: Identity-protective cognition predicts that higher numeracy can worsen accuracy on politically charged facts because cognitive skill is recruited to defend the group line.

4

Explain it step by step

Asking someone to detail how a policy or mechanism actually works destabilizes overconfident belief more than presenting counter-evidence.

Cognitive scientists identified the illusion of explanatory depth: people rate their understanding of everyday objects such as toilets or bicycles far higher than their ability to explain the mechanism. When forced to explain in detail, confidence drops sharply.

Steven Sloman and Philip Fernbach applied the same test to politics. Subjects rated their grasp of a carbon tax or flat tax, then had to walk through how the policy would actually work, step by step. Confidence and often the extremity of the position fell—without any counter-argument.

The technique needs no confrontation. Instead of telling someone they are wrong, you invite them to hear the limits of their own knowledge in their own words. Defensiveness stays low because no group is insulted and no one is attacked.

McRaney treats this as a building block for the later field methods. Both deep canvassing and street epistemology lean on self-generated reflection rather than external argument.

In practice: Subjects who tried to explain a carbon tax or flat tax mechanism by mechanism showed sharp drops in confidence and position extremity with no rebuttal offered.

5

Trade stories, not facts

Lasting attitude change happens when a conversation exchanges personal narrative and vulnerability, letting people update without feeling defeated.

McRaney embeds with the Leadership LAB in Los Angeles, run by Dave Fleischer, which trains canvassers in deep canvassing. The method was refined after the LaCour scandal and validated by the honest Broockman and Kalla replication.

Concrete steps: ask an open, nonjudgmental question about the person’s view; listen without interrupting or correcting; invite a personal memory tied to the underlying value—such as a time they felt discriminated against—then ask them to connect that memory to the group or issue; share a vulnerable story of your own. Canvassers do not argue or cite discrimination statistics.

Field trials produced measurable, durable shifts in support for transgender rights and immigration attitudes that lasted months after a single conversation—effects larger and more persistent than mailers, ads, or standard door scripts.

The proposed mechanism is that the person reaches the new conclusion through their own values and memories, so the change feels like growth rather than surrender. Emotion and story sit at the center on purpose.

In practice: Leadership LAB deep-canvass conversations produced months-long shifts on transgender rights and immigration after one reciprocal story exchange rather than a statistics dump.

6

Score the confidence

Calm Socratic questions that make someone rate their certainty and inspect its method can loosen deep convictions without challenging the belief’s content.

McRaney turns to Anthony Magnabosco and the street-epistemology movement, inspired by philosopher Peter Boghossian. Practitioners use structured questioning in public—often filmed—on religious, political, and paranormal claims.

The method never states a counter-position. The believer rates confidence on a numerical scale, explains what method or evidence produced that number, then considers whether the same method would reliably yield true beliefs in other domains. The goal is to expose the scaffolding under certainty, often for the first time in a nonhostile setting.

Interviews describe people who later abandoned firm beliefs, including religious conviction, not because they were shown wrong but because their own reasoning looked thinner than expected. On-the-spot conversion is rare; more often a question unsettles confidence over weeks or months alongside other life events.

The shared feature with deep canvassing is that the persuader stops being an adversary and becomes a collaborator in the other person’s inquiry, which blocks the identity-defense reaction.

In practice: Street-epistemology conversations ask believers to give a 1–10 confidence score and test whether their method of knowing would work in unrelated domains, often planting doubts that surface later.

7

Exit without total loss

People leave conspiracies and extremist groups through accumulated doubts, a trusted relationship, and a moment that makes leaving feel possible rather than humiliating.

McRaney recounts Charlie Veitch, a prominent flat-earth and 9/11 conspiracy figure who publicly renounced his views after the 2018 Flat Earth International Conference, where a scientist took his questions seriously instead of mocking them. He also draws on cult-exit work associated with Steven Hassan, a former Moonie who developed the BITE model of undue influence.

Across cases the pattern is consistent. Belief systems that feel airtight from inside are held partly by the social cost of leaving—loss of community, status, and identity. Exit therefore needs intellectual doubt plus an alternative source of belonging, or at least assurance that departure will not mean total isolation.

Motivational interviewing, developed for addiction treatment by William Miller and Stephen Rollnick, supplies a practical stance: avoid telling the person what to do; help them articulate their own ambivalence and their own reasons for change.

Conspiracy belief and ordinary political disagreement sit on the same continuum, differing in degree of social entrenchment rather than in kind. The same conversational tools therefore apply to both.

In practice: Charlie Veitch left flat-earth circles after a 2018 conference scientist answered his questions without mockery, inside a setting that did not demand instant public humiliation.

8

Curiosity is the solvent

Genuine, non-strategic curiosity about why someone believes what they believe is the one consistent ingredient across every technique that works.

In the closing synthesis McRaney argues that deep canvassing, street epistemology, and exit counseling all succeed by asking real questions and actually listening, rather than using questions as setups for rebuttal. People detect the difference between a question meant to understand and one meant to trap; the latter reliably triggers identity defense.

He returns to the LaCour episode for a final irony: the fraudulent study got the broad conclusion right—canvassing conversations can work—yet only the honest, rigorous rebuild showed the real effect was smaller, slower, and more procedural than the original headline.

The desire for a quick satisfying technique is itself a trap. Real persuasion requires patience, humility about one’s own certainty, and willingness to be changed by the conversation. Deep-canvass talks take roughly twenty minutes; street-epistemology effects may take months; exit counseling can take years.

Anyone hunting a fast rhetorical trick to win arguments has misunderstood the project. The research shows that the appearance of trying to win is exactly what keeps minds from changing.

In practice: McRaney notes that canvass conversations, confidence-rating dialogues, and multi-year exit counseling share one detectable feature: questions asked to understand rather than to corner.