The Dunning-Kruger Effect: How Incompetence Fuels Overconfidence

The Dunning-Kruger effect is a judgment bias identified in 1999 by psychologists David Dunning and Justin Kruger of Cornell University, and published in the Journal of Personality and Social Psychology. Its central mechanism: people with little skill in a given area paradoxically overestimate their competence[1], precisely because they lack the metacognitive abilities needed to assess the gap between their actual level and the expertise expected.

In other words, the skill required to excel in a field is the same skill required to recognize one's own shortcomings. Without it, self-assessment remains blind.

Online, it is often represented by a three-stage confidence/competence curve, which does not appear in the 1999 study:

• The early peak of overconfidence: the novice picks up a little knowledge and believes they are an expert.
• The valley of humility: as they progress, they realize how much they still do not know.
• The plateau of consolidation: the expert gains a calibrated confidence, based on real mastery.

This curve remains a striking image, not a scientific result.

It all begins in 1995, in Pittsburgh. A man named McArthur Wheeler decides to rob two banks without a disguise, convinced that lemon juice would make his face invisible to surveillance cameras, like invisible ink. Arrested that same evening thanks to the footage, he is stunned when shown the evidence. His incompetence was total. His confidence, absolute.

This case directly inspires David Dunning and Justin Kruger, who design a series of experiments at Cornell University. Their protocol: have participants take skills tests covering humor, grammar and logical reasoning, then ask them to rate their own performance.

The results are striking. The bottom 25% of performers, whose scores placed them at the 12th percentile, estimated themselves on average at the 62nd[1], a clear overestimation of their real abilities. Conversely, the most competent participants tended to slightly underestimate their level, assuming that others had obtained comparable results.

Published in 1999, this study offers an explanation that has remained famous: incompetence does not merely produce poor performance, it also deprives people of the metacognitive ability to recognize it[1].

A debated effect. The interpretation is contested. Some researchers argue that much of the famous curve is a statistical artifact rather than a lack of lucidity[2]. So keep the idea, which is useful in everyday life, without treating it as a law of psychology.

Three distinct mechanisms explain why incompetence prevents people from recognizing incompetence. They feed one another and form a cycle that is hard to break.

The absence of metacognitive skills

Judging the quality of your own performance requires exactly the same skills as those being evaluated. Someone who does not master a field cannot identify their mistakes or measure the gap with real expertise.

Psychologist Daniel Kahneman sheds light on this mechanism with his distinction between two modes of information processing: System 1, made of intuitions and automatic reflexes, fast but prone to distortions, and System 2, analytical and rational, but costly in cognitive effort. Overconfidence often arises from a System 1 that System 2 has not corrected.

The illusion of transparency and the false consensus effect

People with little skill project their own level of knowledge onto those around them. If something seems simple to them, they assume it is simple for everyone. This cognitive projection leads them to underestimate the real complexity of the field and to overestimate the share of it they master. The illusion of competence is thus fed by ignorance of what they do not know.

The anchoring effect on first impressions

Once a person has formed an opinion of their own level, often overestimated from the start, they resist contradictory feedback. The anchoring effect refers to our tendency to stay with the first impression or the first idea we had about a subject. This confirmation bias filters later information to make it compatible with the initial belief, making any external correction particularly hard to absorb.

These two concepts are often confused. Yet the confusion is avoidable.

The Dunning-Kruger effect affects people whose skill is low and who see themselves as experts. Any successes are attributed to their own qualities, while failures are ignored or played down. The overconfidence is unconscious: it is not deliberate arrogance, but a cognitive blind spot.

Impostor syndrome works the other way around. Described as early as 1978 by psychologists Pauline Clance and Suzanne Imes, it affects genuinely competent people who see themselves as impostors. They attribute their successes to luck or circumstances, and their failures to deep personal shortcomings. The constant fear of being "found out" generates chronic anxiety, regardless of objective results.

The two phenomena can coexist in the same person depending on the context: an overconfident novice in a field they have just discovered, an expert who underrates themselves in an adjacent field where they lack reference points. This alternation is normal. It is a reminder that neither confidence nor doubt is a reliable indicator of real competence.

The phenomenon does not stay in the laboratory. You probably run into it every week.

• In the workplace: a candidate comes to an interview with an impressive self-assurance, but their answers reveal gaping holes. A recently promoted manager imposes decisions without consulting the team, convinced they have mastered a field they have only just joined. These situations create tension, weaken team cohesion and harm the quality of deliverables. The incompetence in the workplace linked to this bias is not a matter of ill will: the person concerned simply does not perceive their own blind spots.

• On social media: ultracrepidarianism, meaning giving one's opinion outside one's area of expertise, flourishes there unchecked. Self-proclaimed experts in health, economics or sports strategy broadcast analyses with a confidence inversely proportional to their real training. The lack of corrective feedback amplifies the phenomenon: faced with a subject they know poorly, the audience may mistake assurance for competence.

• In personal life: a driver convinced they are above average behind the wheel, a parent certain they hold the right parenting method without ever having questioned it. These everyday certainties are often harmless, but they show how far this cognitive bias reaches into every area of life.

The 1999 article sets out four predictions about people with low competence. These are not stages that someone goes through, unlike the "Mount Stupid" curve that circulates online and does not appear in the study.

Overestimating one's level

Without a reliable internal frame of reference, a person with low competence judges their level based on what they know, without measuring what they do not know. The lack of objective benchmarks produces excessive confidence, unrelated to actual performance.

Failing to recognize other people's expertise

Watching an expert at work is not enough to grasp the depth of their mastery. The affected person lacks the cognitive tools to tell apart what comes from built-up skill and what simply seems "logical" to them.

Failing to see one's own limits

Without internal feedback and while resisting external feedback, the person stays locked in their initial perception. The vicious circle closes: the less you know, the less you know that you do not know.

Becoming aware of one's limits through training

Acquiring real skills changes everything. In the study, improving participants' skills helped them recognize their limits[1]. Structured training therefore remains a concrete way to adjust one's confidence.

How can you get past the Dunning-Kruger effect?

One clarification first: there is no test for the Dunning-Kruger effect. It was observed by comparing groups in studies. It is not an individual diagnosis, and an online quiz will not tell you whether you are affected.

You can, however, check your own judgment. Ask for precise feedback from people who master the field, compare your results with objective criteria rather than with your impression, and write down your predictions so you can compare them with the facts. When a subject seems simple to you after a few readings, ask yourself what you do not see yet.

If you manage a team, distinguish confidence from competence when hiring, base evaluations on facts, and make regular feedback a habit rather than a sanction.

Mental health professions are not exempt: adopting a new tool requires supervised practice. This is why C2Care supports practitioners who integrate virtual reality into their practice with training and ongoing support.

Key takeaways on the Dunning-Kruger effect

The Dunning-Kruger effect describes a tendency: the least competent people in a field often overestimate their level[1]. It comes from a study published in 1999, and its interpretation is still debated[2]. It can affect anyone, in the areas where they lack reference points. Neither confidence nor doubt guarantees competence: outside feedback, objective criteria and training remain the best ways to adjust one's judgment.