You know the type. The colleague who corrects your numbers in front of the whole team, wrongly. The uncle who has an answer for everything, including things he heard about yesterday. The friend who never once, in ten years, has said “fair point.”
Talking to them is tiring. You walk away replaying the conversation, thinking of the comeback you should have used, and wondering whether you’re the unreasonable one.
Probably not. But you also can’t make them change, and trying usually makes things worse. What you can change is how you handle the conversation, and how much of yourself you leave behind in it. If you’d like a refresher on the basics first, our guide on how to have a good conversation with anyone covers the groundwork. Below are 16 ways to deal with people who always think they’re right, each tied to research or expert practice you can check.
1. Understand why they’re so sure
People who always think they’re right aren’t usually lying to you. Many of them are genuinely, almost serenely, confident. And that confidence tends to travel together with a distorted sense of who agrees with them.
Gordon Pennycook and Jabin Binnendyk at Cornell, working with David Rand at MIT, ran eight studies with 4,181 American adults and published them in Personality and Social Psychology Bulletin in May 2025. People who believed false conspiracy claims consistently overrated how well they’d done on numeracy and perception tests. They also thought they were in the majority about 93% of the time, even though in every study they were a minority. The research focused on conspiracy beliefs, so it doesn’t describe every stubborn person you know. Still, the pattern is familiar: overconfidence in your own judgment and a belief that “everyone thinks this” feed each other.
A second study shows why it matters for you. Han Lin and Yonghwan Kim surveyed 1,175 South Koreans, then 948 of them again a month later, for a 2025 paper in Communication Research. Talking with political opponents led to more hostile behaviour, and eventually more polarization, only among people who overestimated their own knowledge.
So with an overconfident person, a head-on clash tends to harden them. Start small: next time they dig in, quietly label it to yourself (“this is certainty talking, not evidence”) before you say anything at all.
2. Decide whether this conversation is worth having
You don’t owe anyone a debate. Plenty of people have already worked this out.
In a survey of 3,560 U.S. adults fielded in December 2025, Pew Research Center’s February 2026 report on why Americans discuss the news, or don’t, found that 56% had stopped talking to someone about political or election news because of something that person said. That’s up from 45% in 2024. The most common reason people gave for avoiding news conversations was worry about making things uncomfortable (58%).
Some of that avoidance is a loss, sure. But it also shows that stepping out of a pointless argument is a normal, widely used choice, not a failure of nerve. Before you engage, ask yourself three quick things. Does this matter to me? Is anything going to change? Do I have the energy right now? If the answer to all three is no, nod, say “interesting,” and talk about the weather. If the person tends to get heated, our article on how to deal with angry people has more on keeping it from boiling over.
3. Listen like you actually want to hear it
This sounds soft. The data says otherwise.
Dvori Saluk, Guy Itzchakov, Netta Weinstein and Moty Amar ran five studies with 1,624 people on what happens when someone listens well during a disagreement. Their paper, Listening Across the Divide, in press at the Journal of Personality and Social Psychology (2026), included live Zoom conversations. When the listener paid close attention, speakers reported much more positive feelings (Cohen’s d = 0.91) and a stronger sense of connection (d = 0.90) than when the listening was only moderate. In a field survey of real disagreements, conversations where people felt more heard than usual came with less negative emotion too.
One honest limit: the study measured how the speaker felt, not whether they changed their mind. But a person who feels heard is calmer, and calm people are easier to deal with. Listening well also isn’t the same as waiting for your turn. If you catch yourself steering every topic back to your own view, our guide on how to stop talking about yourself in conversations is a useful companion.
Test it this week: in one conversation with your know-it-all, don’t argue for the first two minutes. Just ask follow-up questions.
4. Reflect back what you heard
Seth Izen runs MetroWest Mediation Services, a Massachusetts nonprofit that has trained mediators since 1979. Writing for WBUR in May 2026 about families split by politics, he suggests that you stop trying to persuade and aim to understand instead. Success, in his view, means keeping the relationship, not winning the point.
His main tool is reflecting. You restate what the other person said in your own words, without signalling agreement. Something like “What I’m hearing is that you think the new policy wastes money.” That’s it.
Why bother? Because people who insist they’re right often keep repeating themselves since they don’t believe you’ve understood. Reflecting closes that loop. It also slows the conversation down, which helps you as much as them. Izen also points out that people tend to mirror the tone they receive. Bring calm curiosity, and you raise the odds of getting some back. For a broader set of skills along these lines, see how to improve social skills.
Your move: the next time they finish a rant, summarize it in one sentence and ask, “Did I get that right?”
5. Assume they’ve only heard half the story
There’s a generous way to think about stubborn people, and it happens to be backed by evidence. They may not be stupid or arrogant. They may simply not know what they don’t know.
Hunter Gehlbach (Johns Hopkins), Carly Robinson (Stanford) and Angus Fletcher (Ohio State) tested this in a study of 1,261 Americans, published in PLOS ONE in October 2024. Everyone read about a fictional school with a water problem. Some read arguments for only one solution, while a control group read both sides. The one-sided readers felt they had enough information. They were more confident than the people who had the full picture, and they assumed most others would choose the same thing. “They were quite sure that their decision was the right one,” Fletcher said. Encouragingly, some changed their minds once they read the other side.
There’s a catch, and Fletcher named it himself. On deeply ideological issues, people may distrust new information or spin it. So this works best on practical disagreements: which supplier to use, the best route to the airport, how to load the dishwasher.
Instead of saying “you’re wrong,” offer what’s missing: “Have you seen the part about the cost?” It lands softer, and it gives them an exit.
6. Calm your own body before you open your mouth
Dealing with people who always think they’re right is stressful, and that stress shows up physically.
The American Psychological Association’s Stress in America 2025 survey, run by The Harris Poll with 3,199 U.S. adults in August 2025, found that 62% named societal division as a significant source of stress. Among those people, 83% reported at least one physical stress symptom in the past month, compared with 66% of everyone else. That’s a correlation, not proof that arguments cause headaches. But it matches what most people feel after a tense exchange: tight jaw, racing heart, the urge to snap.
And a body in that state argues badly. You interrupt more, you hear less, and you say things you’ll regret. If you want a few daily habits for staying steady, our Stoic exercises for daily calm are a good place to start.
Next time, do this: before you reply to something infuriating, breathe out slowly for longer than you breathed in. Twice. Then speak.
7. Acknowledge their point before you add yours
Julia Minson, a behavioural scientist and professor of public policy at Harvard Kennedy School, has spent years studying how people disagree. In a March 2026 Harvard Magazine interview, ahead of her book How to Disagree Better, she laid out a framework she calls H.E.A.R.:
- Hedge your claims (“most of the time,” “in my experience”).
- Emphasize where you agree.
- Acknowledge their view before you disagree.
- Reframe things positively, with fewer negatives.
She also flags words that quietly make things worse. “Never” and “terrible” escalate. “Just” and “simply” sound dismissive. According to Minson, short training in this kind of phrasing makes people come across as more trustworthy and objective to those who disagree with them, and she argues that teaching specific phrasing works better than simply urging people to be more empathetic.
The beauty of this approach is that it doesn’t ask you to agree. It asks you to show you understood before you push back. That’s an emotional intelligence skill as much as a language one, and our guide on how to improve your EQ goes deeper on it.
Here’s a simple way in: open your next disagreement with “I can see why you’d think that, because…” and finish the sentence honestly.
8. Say what you’re against instead of what you’re for
This one is surprising. How you phrase the first sentence of your opinion changes how open the other person is to it, even if the argument after it is identical.
Rhia Catapano (University of Toronto) and Zakary Tormala (Stanford) ran a series of experiments with more than 10,000 participants, published online in the Journal of Personality and Social Psychology in December 2025. People sharing opinions on topics like abortion and gun control expected that “support” framing (“I support allowing abortions”) would persuade best. In fact, receivers were more open to “opposition” framing (“I oppose abortion bans”). In a simulated Reddit feed, people were also more likely to click on opposition-framed posts.
Why might that be? One plausible reading is that saying what you oppose sounds narrower, almost like a shared concern, while “I support X” can sound like you’re asking them to sign up to your whole worldview. The researchers didn’t pin down the mechanism, so treat that as an educated guess.
A quick experiment: take the point you most want your stubborn colleague to hear, and rewrite the first sentence as something you’re against. “I’m against shipping before testing” instead of “I support more testing.” See if the response changes.
9. Give them a way to back down without losing face
Think about what it costs someone like this to say “I was wrong.” For a person whose identity is built on being the one who knows, a public reversal is a status loss. That’s the mechanism: the more an admission threatens how someone sees themselves, and the more people are watching, the harder they’ll defend the original position, regardless of the facts. That’s a reasoned account built on how status and identity work, not a single tested law, but the next study points the same way. Corner them, and you’ve turned a question about facts into a question about pride.
A popular trick shows how this can go wrong. You may have heard that asking someone to explain, step by step, how a policy works will humble them. A 2013 study by Philip Fernbach and colleagues found exactly that. But when Richard Walker, Jiasheng Zhang and Edmund Cheng repeated the test with 5,139 people in Hong Kong (Political Psychology, June 2024), it backfired: people’s positions became more extreme after they explained their understanding. A gotcha question, it turns out, can feel like an ambush.
Say you’re in a meeting and your manager is confidently wrong about a deadline. Saying “That’s not right” in front of everyone puts them on the spot. Saying “I might be looking at an older version, can we check the latest timeline together after this?” lets them update without anyone keeping score.
Make the exit easy, private if possible, and let them take some credit for spotting the fix.
10. Admit when you’re wrong, out loud
It feels risky to concede anything to someone who never concedes. The research suggests it helps you more than it hurts.
Adam Fetterman, a psychologist at the University of Houston, ran five experiments on this, published in Public Understanding of Science in December 2025. In one, 500 participants read about a scientist whose earlier finding failed to replicate. The scientist either admitted being wrong, insisted the effect was real, or said nothing. Those who admitted the error were rated as more trustworthy, with higher scores for expertise, integrity and goodwill. A separate study by Fetterman found that students rated professors who seemed willing to admit mistakes as both warmer and more competent.
So when you concede a small point (“You’re right, I had the date wrong”), you build credibility that makes your bigger points harder to dismiss. And you’re showing, without lecturing, that being wrong isn’t the end of the world.
Give it a go: find one small, genuine thing to concede in your next disagreement, and say it clearly before you make your main point.
11. At work, use the right channel
Workplace know-it-alls are common, and they wear people down. In November 2025, the UK’s conciliation service Acas published a survey of 4,558 working-age adults, carried out by the National Centre for Social Research in August and September 2025. The headline: 44% had experienced conflict at work in the previous 12 months. Among them, personal disagreements and relationship issues accounted for a third (33%) of cases, and the other party was most often a colleague (39%) or a line manager (32%).
What people did about it is useful to know:
| Response to conflict at work | Share of those in conflict |
|---|---|
| Talked to their own manager | 45% |
| Had an informal chat with the other person | 30% |
| Did nothing | 19% |
| Raised a formal grievance | 9% |
Half said the conflict ended up fully or largely resolved (27% fully, 23% largely).
The table shows that informal routes are by far the most common. Before anything formal, try a calm, private conversation, then your manager if that goes nowhere. Keep notes on specific incidents (“On 4 March, he overrode the agreed figures in the client deck”) rather than character judgements (“he’s arrogant”). If you’re the manager dealing with someone like this on your team, see our guide on how to be a good boss.
Your next step: write down the last three times this person’s certainty caused a real problem, with dates. That list is what you bring to a manager, not a feeling.
12. Bring specific evidence, not a lecture
There’s a cynical idea that facts never change anyone’s mind. The best recent research says that’s too pessimistic, as long as the facts answer the person’s actual reasons.
Thomas Costello (now at Carnegie Mellon), Gordon Pennycook and David Rand had people who believed a conspiracy theory chat with an AI chatbot that responded to their specific arguments with evidence. Published in Science in September 2024, the study found that belief dropped by about 20% on average and stayed nearly as low two months later. A 2025 follow-up in PNAS Nexus found it worked whether people were told they were talking to an AI or a human. A further preprint (not yet peer-reviewed) found that participants who were persuaded mostly credited the evidence-focused approach, not friendliness.
You should know one caveat. In June 2026, Science added an expression of concern to the original paper after the authors themselves reported problems reproducing some values from the public dataset. The authors say their reanalysis doesn’t change the conclusions, and political scientist Brendan Nyhan reported that his Dartmouth team reproduced the results in two separate experiments. The journal is still reviewing it.
The practical lesson holds either way. A generic rebuttal bounces off. A patient, specific answer to the exact reason they gave has a fighting chance. This matters online too, where people double down fast; our guide on how to deal with haters on social media covers that arena.
Start small: ask “What’s the main reason you think that?”, then answer only that reason, with one solid source.
13. Look for the shared value underneath
Braver Angels is a U.S. nonprofit that runs structured workshops where, typically, eight Democrats and eight Republicans spend a full day together with trained moderators. The format borrows from couples counselling: each side reflects on its own stereotypes and listens to how the other side sees itself.
The idea has been tested properly. A randomized trial with undergraduates at four universities, summarized by Brown University’s Watson Institute in January 2025, found the workshops substantially reduced polarization on both explicit and implicit measures, though some effects faded over time. Then Lee-Or Ankori-Karlinsky, Robert Blair, Jessica Gottlieb and Samantha Moore-Berg tested whether simply watching a 50-minute documentary about one of these workshops could help. Their randomized trial with a nationally representative YouGov sample, published in Political Science Research and Methods in August 2025, found it reduced affective polarization by 0.14 standard deviations about five days later. By roughly 50 days, that effect had shrunk and was no longer statistically significant. But viewers’ belief that dialogue can change things held up.
That’s a modest, honest result. Shared ground helps, and it needs topping up. In everyday terms, the policy you’re arguing about often sits on top of values you share, like safety, fairness or wanting the kids to be OK. If you can name that value first, you’re arguing as two people with the same goal and different routes, rather than as opponents.
Before the next round with your relative, figure out what they’re actually worried about, and say that part out loud.
14. Stop pulling on the rope
Arguing with someone who always has to be right is a lot like a tug of war. The harder you pull, the harder they pull back, and neither of you moves. Drop your end, and there’s nothing left to pull against.
The comparison holds in one important way: resistance feeds on resistance. It breaks down in another, because people aren’t ropes, and dropping your end doesn’t mean they’ll suddenly walk over to your side. That’s worth knowing so your expectations stay realistic.
The Pew Research Center survey from February 2026 gives a useful reality check here. Among U.S. adults, 59% said they at least sometimes change their own opinion when discussing the news. Only 4% said that happens all or most of the time. So minds do move, but slowly, rarely in one conversation, and almost never while someone is being pushed.
When you notice you’re both repeating yourselves, that’s the rope. Your move: say “I don’t think either of us is going to convince the other today, and that’s fine,” and mean it.
15. Protect your energy and set limits
Some people who always think they’re right are merely irritating. Others drain you. The second group needs boundaries, not better arguments.
The Acas survey from November 2025 shows what’s at stake when conflict drags on. Among workers who’d been in conflict, 57% reported stress, anxiety or depression as a result, and 49% said their motivation or commitment dropped. Only 17% said it had no effect on them at all. Those numbers cover all kinds of workplace conflict, not only know-it-alls, but they show that ongoing friction usually costs you something.
A boundary is simply a clear statement of what you will and won’t do. “I’m happy to talk about this, but not if you’re going to talk over me.” “I’m not discussing my parenting choices.” You don’t need to justify it at length. If someone repeatedly dismisses your feelings, rewrites what happened, or makes every disagreement about your character, that goes beyond stubbornness. Our article on dealing with a narcissistic person in a relationship may help you tell the difference.
Test it this week: pick one recurring topic you’re done arguing about, write your boundary sentence in advance, and use it word for word the next time it comes up.
16. Hit pause before things escalate
Heated arguments run on a feedback loop. One sharp comment raises your heart rate, which narrows your attention, which makes your next reply sharper, which provokes theirs. Each step pushes the next one along. The research below shows the matching-and-retaliation part directly; the physiology is the likely driver rather than something that study measured. Interrupting the loop, even briefly, gives your thinking a chance to catch up with your feelings.
Researchers at the University of St Andrews tested this with 81 couples in a study published in Communications Psychology in August 2024. Couples played a competitive game in which they could blast each other with unpleasant noise. Partners tended to match each other’s aggression, but enforced breaks broke that cycle. Five-second pauses worked as well as ten- or fifteen-second ones. The researchers stress that this applies to everyday conflict, not abusive relationships, and it was a lab game, not a real argument about money or in-laws.
Still, the principle travels well. A pause can be tiny: a sip of water, a slow “hmm,” or “Let me think about that.” It can also be longer: “I want to come back to this tomorrow when I’m less wound up.” And if the conversation has already ruined your mood, our guide on how to turn a bad day around has some ways to reset.
Next time you feel the heat rise, count five seconds before you answer.
What to do when you slip
You will lose your temper with this person at some point. You’ll take the bait, raise your voice, or deliver a perfectly crafted put-down you instantly regret. Everyone does.
First, don’t let one bad exchange become a story about who you are. If you keep replaying it at 2 a.m., our tips on how to stop ruminating thoughts can help break the loop.
Second, go back and repair it. Use the admission from item 10: “I got heated yesterday, and I’m sorry for that. I’d still like to understand your view.” It costs you very little and often resets the whole dynamic.
Third, remember that disagreement itself isn’t the problem. In one of the studies in the 2026 paper by Saluk, Itzchakov and colleagues, people who disagreed but were listened to well came away feeling more positive than people who agreed but were only listened to moderately. How the conversation goes matters more than whether you see eye to eye. That’s good news, because it means you can recover from a bad round. Next time, listen better.
The bottom line
You probably won’t turn someone who always thinks they’re right into a person who says “good point” twice a day. That’s not the goal. The goal is to stay calm, stay curious, stop wasting energy on arguments that go nowhere, and protect the relationships worth keeping.
Pick two strategies from this list, the ones that felt most doable, and use them in your next tricky conversation. Then come back for more practical advice on handling the people in your life in our Relationships section.
