How Do You Challenge What You Can’t Immediately Disprove?

Recently, I read an email from The Atlantic that took me behind the scenes of its fact-checking process, with research chief Yvonne Rolzhausen explaining how far fact-checkers go before a feature is published.

Depending on the story’s complexity, the process can take hours, weeks or even months. The article is reformatted with wide margins. Checkable facts are marked in red pencil, while material with legal implications gets scrutinised, sources are contacted, and claims are traced back to their origins¹.

It made me think about a very different environment: the meeting room and how information is fact-checked (or not) in real time.

Leaders spend much of their working day making decisions based on information supplied by other people. Someone tells you a project will be six months late. A colleague says a competitor is about to make a particular move. You’re told the team can’t absorb any more work, that customers are demanding a certain feature, or that “everyone” is unhappy with a decision.

Sometimes the evidence is clear; other times, claims are delivered confidently, but something about them doesn’t quite stack up.

You might suspect the assertion is exaggerated, incomplete or perhaps just wrong. But you don’t have a better number sitting in front of you, you can’t immediately produce evidence to contradict it, and the conversation is already moving on.

So, what do you do? Say nothing? Challenge based on gut feeling? Something else?

How you respond is likely to be influenced by how confidently the message was delivered, and there are good reasons for that.

Your A–Z guide to navigating the everyday challenges of work →

Why Confident Claims Can Be So Persuasive

In a series of experiments, psychologist Daniel Gilbert and colleagues examined how people process information they are explicitly told is false.

Across several experiments, participants were exposed to information about another person that was identified as false. When they were under time pressure or had another task competing for their attention, they were more likely to later rely on that false information when making a judgement call².

The researchers argued that comprehension initially involves representing information as true, whereas rejecting it as false requires an additional cognitive process that limited processing capacity can disrupt.

That matters because meetings are rarely environments of perfect concentration. You’re listening, thinking about the next agenda item or your next meeting, responding to messages, watching the clock and trying to make a decision, often all at once.

There’s another complication.

Research led by Professor Cameron Anderson suggests that overconfidence can contribute to higher social status. Across six studies, overconfident individuals tended to attain greater status, while one study indicated that overconfidence can produce behavioural cues that make a person appear competent³. The broader finding is not that confidence always signals competence, but that confident behaviour can influence how others evaluate a person, even when confidence is poorly calibrated to the person’s actual ability.

What this means is that, at times, we can mistake confidence for competence.

You’ll have likely seen this play out at work. When someone speaks with certainty, you’re more likely to assume they know what they’re talking about, even when their confidence isn’t matched by the quality of their evidence.

And that’s where the risk lies. A confidently delivered claim can carry more weight.

You Don’t Need a Counterclaim To Ask For Evidence

One of the key issues we face is assuming that, before we challenge something, we need to know the alternative.

If someone says the number is 70 per cent, we feel we need another number. If they say the project will fail, we feel we need to prove that it won’t. If they say, “Everyone thinks this”, we feel we need to know what everyone thinks.

But challenging a claim isn’t the same as contradicting it. You don’t have to say, “That’s wrong.” You can say, “Help me understand what that’s based on.”

The first invites a debate about who’s right. In contrast, the second shifts the conversation to the evidence.

Read more: Holding Your Ethical Line: When Voicing Concerns Isn't Enough

Start By Tracing The Claim

So, if you’re in a meeting and you want to fact-check a claim, where do you start?

One of the simplest questions you can ask is, where did that come from? It sounds almost too obvious, but it improves the conversation because it focuses it on the source.

For example:

• Was the figure measured, modelled, estimated or remembered?
• Is it based on customer data, three conversations, a survey or someone’s impression?
• Did the competitor announce the change, or is the information third-hand?
• When someone says, “The team thinks”, who exactly has been asked?

The key is to make this kind of questioning part of how you work, rather than something you only do when you disagree.

If you only interrogate evidence when you disagree with the conclusion, people will quickly learn that asking for sources is really a proxy for challenging the person.

Consequently, a better discipline is to apply the same curiosity to the claims you agree with as the ones you don’t. That way, you demonstrate that your questioning isn’t about whether the evidence suits you, but about whether the evidence is strong enough to support the decision.

Ask How Certain They Really Are

The next step is to separate confidence in delivery from confidence in the evidence.

Try asking: How confident are you in that?

And then: What would change your mind?

I particularly like the second question because it shifts the discussion away from defending a position to examining the limits of knowledge and confidence. Even well-founded conclusions have limits and assumptions.

Work Out Whether The Claim Matters

Not every questionable statement deserves a forensic investigation. Leaders don’t have the time, and organisations would grind to a halt if every decision went through The Atlantic’s fact-checking process.

So the next question is: What would we do differently if this turned out to be wrong?

If the answer is “nothing”, there may be little value in pursuing it. But if the answer is “we would make a completely different decision”, then you’ve identified something that deserves more scrutiny.

This is where verification becomes a judgement issue rather than an administrative one. The claims worth checking most carefully aren’t necessarily the easiest ones to check, but the ones carrying the biggest risk and consequence.

For example, if your investment decision depends on market size, test the market assumption. If the restructure depends on a claimed capability gap, examine the evidence for that gap. If a decision affecting an employee is based on what “people are saying”, find out who said what and in what context. If launching a new product depends on an emerging consumer trend, investigate where the data comes from, how strong the evidence is and whether the trend has staying power.

Don’t Let “We’ll Check Later” Be The End

Research on misinformation describes what is known as the continued influence effect. Even after inaccurate information has been corrected, it can continue to influence the way people think and reason⁴.

That doesn’t mean every mistaken comment in a meeting or conversation becomes permanently embedded. But the findings suggest a practical organisational precaution: when an important claim has influenced a discussion or decision, don’t assume a later correction will automatically remove its effects from what people are thinking.

So if an important claim can’t be verified in the room, don’t simply say, “Let’s take that offline.” Be more specific by asking:

• Who’s going to check it?
• What exactly are they checking?
• What evidence will they bring back?
• And when will the group revisit the decision?

Most importantly, close the loop with the same people who heard the original claim. If the original claim was communicated publicly, the correction should ideally reach the same audience.

Build Verification Into The Way Your Team Works

Of course, the best time to decide how your team handles evidence isn’t when you’re already debating a contentious claim. It’s beforehand, by instituting better practices and habits.

This approach might mean asking people to include sources for critical figures in decision papers. Distinguishing clearly between data, assumptions and estimates. Identifying the key assumptions behind a recommendation. Or explicitly asking, before an important decision, “What are we treating as true here that we haven’t yet tested?”

There’s also a cultural element to factor in: making it psychologically safe for your team members and colleagues to challenge. Questioning another person’s claim can be (and feel) risky, particularly when that person is more senior, confident or influential.

Consequently, pay attention to what happens after someone asks the uncomfortable question. Do you get curious or irritated? Do you explore the challenge, or do you subtly signal that it would be easier if people stopped asking?

You can tell people you want challenge, candour and rigorous debate, and yet, your reaction in those moments will tell them whether you really mean it.

Create The Space To Test What Matters

You’ll rarely have perfect information, so the goal isn’t to verify everything. It’s to recognise when the evidence matters enough to pause before deciding.

When a confident claim lands in front of you (whether it’s in a meeting or conversation) and something about it doesn’t feel right, you don’t need to have the rebuttal ready. You can simply ask:

• Where did that come from?
• What would change your mind?
• What would we do differently if it were wrong?
• What do we need to verify before we decide?

Being sceptical doesn’t mean you have to prove the opposite on the spot. If a claim is important enough to influence the decision, it should be strong enough to withstand scrutiny.

So, as you look at the decisions in front of you, ask yourself: What are we treating as fact that nobody has actually verified?

Republished with courtesy from michellegibbings.com.