The Confidence Calibration Problem – Why Self-Assessment Fails Us at the Worst Possible Moments
The Confidence Calibration Problem
Why Self-Assessment Fails Us at the Worst Possible Moments
Issue 247, January 15, 2026
In virtually every meeting where decisions are being made, someone in the room is speaking with authority on a subject they understand poorly, while someone else with genuine expertise is holding back because they’re not confident their perspective or even the information they have is valuable. This asymmetry shapes outcomes in ways we rarely recognize.
This phenomenon of human behavior cuts across all decision types, whether they be significant change and transformation-related or the smallest operational and business decisions. We all face this in one way or another; perhaps we are the ones holding back, or perhaps we are the one who feels we must speak with authority, even though we know very little about the subject.
The research on this topic, published by Kruger and Dunning in the Journal of Personality and Social Psychology, has been replicated so consistently that it’s become a cornerstone of behavioral psychology. In their original 1999 study, people scoring in the bottom quartile of performance estimated themselves at the 62nd percentile when they were actually at the 12th. In other words, people with the least knowledge or skill in a given domain systematically overestimate their competence, while those with the greatest expertise tend to underestimate theirs. The gap between actual ability and self-assessed ability isn’t random; it follows a predictable pattern everywhere that judgment matters.
The Mechanics of Miscalibration
The core insight is deceptively simple: the skills required to produce excellent work are the same skills required to recognize excellent work. Someone who lacks a genuine understanding of a domain also lacks the ability to assess their own understanding accurately. Their confidence isn’t evidence of competence; it’s evidence that they can’t see what they’re missing. Kruger and Dunning termed this the “dual burden” of incompetence, where poor performers suffer twice: first in making errors, and second in being unable to recognize those errors.
Subsequent research has confirmed and extended these findings. Work by Pennycook and colleagues, published in Psychonomic Bulletin & Review, found that participants making the most errors on a cognitive reflection test overestimated their performance by a factor of three or more. Meanwhile, experts often recognize how much they still don’t know, which makes them more hesitant to claim mastery. A follow-up study by Dunning and colleagues, published in Organizational Behavior and Human Decision Processes, documented that bottom-quartile performers believed their mastery was at the 63rd percentile when they were actually at the 15th. The same expertise that enables good judgment also reveals the limits of that judgment.
Importantly, subsequent scholarship has added nuance to these findings. The effect appears strongest in domains where feedback is ambiguous or delayed, such as leadership, strategic planning, and interpersonal communication, precisely the domains where organizational decisions carry the greatest weight. In areas with clear, immediate feedback (like math problems with right or wrong answers), people calibrate more quickly. This suggests that the confidence calibration problem isn’t merely about individual psychology; it’s amplified by organizational structures that often lack mechanisms for timely, unambiguous feedback on complex decisions.
This creates an ironic situation in group settings. The person who speaks most confidently may understand the topic least. The person with the most valuable perspective may hold back, assuming that others must know something they don’t. Decisions get made based on who projects certainty rather than who actually understands the problem. In a nearly constant state of environmental and marketplace change, where precision and critical thinking, with a little luck ensures an organization sustains itself, competes well, or grows, our own very humanness stands once again in our way.
Where This Shows Up
Watch any strategy discussion, and you’ll see this pattern. Someone with surface-level knowledge makes bold assertions about market dynamics, customer behavior, or competitive threats. They speak with conviction because they don’t know what they don’t know. Meanwhile, the person who has studied the market deeply qualifies their statements, acknowledges uncertainty, and sounds less decisive as a result.
The same dynamic appears in technical discussions. The engineer who has thought deeply about system architecture sees the trade-offs and complications. The manager who learned a few buzzwords last week sees only the exciting possibilities. Guess which perspective often sounds more compelling in a meeting?
It appears in hiring decisions, where confident candidates may know less than thoughtful ones. It appears in performance reviews, where self-assessments correlate poorly with actual contribution. It appears in leadership development, where the people who are most certain they’re ready to lead may be the least ready to learn.
Recognizing Miscalibrated Confidence
Once you understand this pattern, you can watch for signals that someone’s confidence exceeds their competence. Certainty without qualification is one warning sign. Experts almost always acknowledge conditions, exceptions, and limitations. People who present complex situations as simple may be revealing how little they’ve thought about them.
Resistance to challenge is another signal. People who deeply understand a subject welcome good questions because they’ve thought about the answers. People whose understanding is shallow often become defensive when questioned because they can’t defend what they don’t actually understand.
Watch for the absence of “I don’t know.” Expertise includes knowing the boundaries of what you know. People who never express uncertainty are either genuinely certain about everything, which is impossible, or unable to recognize what they don’t know, which is a different problem entirely.
Recognizing It in Yourself
The harder challenge is recognizing when your own confidence is miscalibrated. The nature of the problem means you can’t simply introspect your way to accuracy. If you could accurately assess your own competence, you wouldn’t have the problem in the first place.
A few practices help.
First, develop genuine relationships with people who will tell you the truth. Not people who admire you, agree with you, or want something from you. People who respect you enough to disagree with you. Their feedback provides external calibration that self-assessment can’t.
Second, track predictions. When you’re confident about how something will turn out, write it down. Check later whether you were right. Over time, you’ll develop a more accurate sense of when your confidence is warranted and when it’s running ahead of your knowledge.
Third, cultivate genuine curiosity about domains where you feel certain. The more you learn about any complex topic, the more you discover that you didn’t know. If learning isn’t increasing your appreciation of what you don’t understand, you may not be learning as much as you think.
Creating Environments Where Expertise Gets Heard
Leaders can counteract this pattern by designing environments where expertise matters more than confidence. This means deliberately soliciting input from quiet people before confident people have established the frame. It means asking experts to speak first in their domains rather than last. It means valuing thoughtful uncertainty over glib certainty.
It also means building cultures where admitting uncertainty is safe. Remember how often we emphasize the importance of psychological safety in any work culture. The more safety the workforce feels, the more accurate and relevant information flows.
If acknowledging what you don’t know is penalized, people learn to project confidence regardless of whether it’s warranted. If asking questions is rewarded, people develop more accurate self-assessment over time.
The goal isn’t to silence confident people or paralyze decisions with endless qualification. The goal is to ensure that the relationship between confidence and competence in your organization is something closer to what it should be: correlated rather than random or inverted.
Your best decisions will come when the people who know the most feel empowered to share what they know, including what they know they don’t know. Getting there requires understanding why that’s harder than it sounds, and deliberately building environments where expertise rather than assertion carries the day.
When you consider all that today’s organizations face, the more focus we apply to the human factors, the more fruitful our efforts become to achieve the desired goals.
Connect with Us
What leadership challenges are shaping your decisions right now? We’d love to hear your insights. Share your experiences with us on our Substack or join the conversation on our LinkedIn. For more insights on navigating organizational complexity, explore our archive of “Ideas and Innovations” newsletters or pick up a copy of The Truth About Transformation: Leading in the Age of AI, Uncertainty and Human Complexity.
Go Deeper: Subscribe to the Human Factor Podcast
If you haven’t yet subscribed to the Human Factor Podcast, find it on your favorite podcast platform. Over the first 13 episodes, Kevin has covered frameworks and strategies to mitigate and contend/strategize how to change and transform. It’s destination business strategy listening!
Season 2 is coming in mid-February 2026
Listen and view on:
