The Dunning-Kruger Effect: What the Research Actually Showed — and Why It Matters to Entrepreneurs from Two Angles

The Dunning-Kruger effect is one of the most famous concepts in popular psychology. Its essence, as often retold, sounds roughly like this: incompetent people are unaware of their incompetence and therefore overestimate themselves, while the smart, conversely, underestimate themselves. Hence the popular chart featuring a mountain: a novice at the peak of confidence, then a plunge into the 'valley of despair,' followed by a long climb toward actual competence.

The problem is that this chart does not come from the original study. It is not in the 1999 paper. And some of what is attributed to Dunning and Kruger has faced serious statistical criticism in recent years.

This does not mean the effect does not exist. It means it is structured differently than is commonly thought — and it is this correct understanding that makes it a useful tool for the entrepreneur.

What the 1999 study actually showed

Justin Kruger and David Dunning of Cornell University published an article in the Journal of Personality and Social Psychology in 1999 titled "Unskilled and Aware of It: How Difficulties in Recognizing One's Own Incompetence Lead to Inflated Self-Assessments." Dunning received an Ig Nobel Prize in Psychology in 2000 for this and related work.

The study consisted of four experiments. Participants were tested in several areas—logical reasoning, grammar, and humor recognition—and were asked to rate their results relative to others. Participants were divided into quartiles based on their actual performance.

Key findings: Participants from the bottom quartile (actual result roughly at the 12th percentile) estimated themselves to be around the 62nd percentile—an overestimation of 50 percentile points. Participants from the top quartile (actual result around the 86th percentile) estimated themselves to be around the 68th percentile—meaning they slightly underestimated themselves.

The authors proposed a mechanism: the skills required to perform a task well are the same skills needed to evaluate one's own work. An incompetent person suffers a double burden: they perform poorly and cannot recognize it because they lack metacognitive skills—the ability to think about the quality of one's own thinking.

The fourth experiment was illustrative. Participants in the bottom quartile were taught logical reasoning, after which they were asked to re-evaluate their results. After training, their self-assessment significantly decreased—meaning that as they acquired the skill, they also acquired the ability to see how poorly they had performed before.

What was challenged — and what remains valid

Since 2016, serious statistical criticism has emerged. Edward Nuhfer and his colleagues published two papers in the journal Numeracy, showing that the characteristic pattern of the Dunning-Kruger effect can be replicated with random data. The problem lies in the methodology of the original study: when the same variable (test score) is used both to sort participants and to calculate the self-assessment error, it creates a statistical artifact—autocorrelation. Gignac and Zajenkowski showed in 2020 that when using continuous variables instead of quartile division, the effect decreases substantially.

Does this mean the effect does not exist? No. Jansen, Rafferty, and Griffiths published a major replication in Nature Human Behaviour in 2021 with computational modeling and a sample of 4,000 participants. Their conclusion: the basic pattern is replicated, but its mechanism is different from what Dunning and Kruger assumed. The effect is explained not so much by a specific metacognitive deficit in the incompetent as by general 'noise' in self-assessment—all people are inaccurate in their self-evaluations, but this creates different consequences at different ends of the competence scale.

A fair summary of the discussion resulting from the data:

The extreme version of the effect—'the incompetent are wildly overconfident, the competent significantly underestimate themselves'—is exaggerated and partially a statistical artifact. The broad conclusion, which is not disputed: people are generally poorly calibrated in self-assessment, and low competence creates a higher risk of overestimation than high competence. The famous 'mountain of confidence' chart is not from the original study and does not reflect its data.

The flip side: Imposter syndrome

This topic has a mirror effect, which is mentioned less frequently in popular retellings.

In 1978, clinical psychologists Pauline Clance and Suzanne Imes published an article in the journal Psychotherapy: Theory, Research and Practice about what they termed the imposter phenomenon. The work was based on clinical observations of more than 150 successful, high-achieving women—professors, graduate students, and professionals.

Despite objective achievements, these women consistently attributed their success to external factors: luck, an error by the evaluator, or a fortunate sequence of events. They lived in fear of being 'exposed' as insufficiently competent—despite the fact that evidence of their competence was accumulating. Clance and Imes called this the feeling of being an 'intellectual imposter.'

Later studies, including a systematic review by Bravata et al. in the Journal of General Internal Medicine (2020), showed that the imposter phenomenon is much more widespread than the original sample suggests. It occurs in men and women, across different professions and cultures. It is especially common in people with high standards for themselves, those in new or competitive environments, and those with a strong orientation toward results.

Why this concerns entrepreneurs from two sides

It is important not to confuse which effect is at play, as entrepreneurs are vulnerable to both in different situations.

The Dunning-Kruger effect in an entrepreneurial context manifests primarily when entering a new field. A founder who is doing marketing for the first time, evaluating a deal's legal structure for the first time, setting up sales, or entering a foreign market for the first time is precisely in the risk zone for overestimation. This is not because they are stupid, but because they lack metacognitive benchmarks in that specific area: they do not yet know what it is they do not know.

This has direct consequences. 'Unknown unknowns' are the most dangerous category. A known risk can be insured or factored into a plan. A risk you are unaware of does not make it into any calculation. The studies by Dunning and Kruger, even with adjustments for statistical criticism, describe the mechanism well: in the early stages of mastering any area, a person cannot see its complexity in its entirety.

Imposter syndrome in an entrepreneurial context works differently and often affects different people at different times. A founder who has already achieved something significant—secured their first clients, closed a round, grown several times over—often cannot internalize it. They attribute success to market conditions, luck, or the team—to everything except their own contribution. This creates a persistent state of anxiety where the person fears 'exposure' and avoids situations where that exposure might occur: new markets, larger clients, or public visibility.

Both effects are problems of calibration. But they lead to opposite errors. Uncalibrated self-confidence leads to moving into areas where one lacks competence, without preparation or expert support. Uncalibrated self-effacement leads to avoiding steps for which the person is objectively prepared.

How to calibrate self-assessment: What helps

Research points to several mechanisms that help in more accurately assessing one's own competence.

Seek feedback from those who know the subject. In their original article, Kruger and Dunning noted that in everyday life, people very rarely receive honest negative feedback about their skills. Colleagues, clients, and partners often suppress criticism. An entrepreneur without a systematic point for receiving honest feedback—a mentor, an external expert, or a structured review of results—operates in an informational vacuum where self-assessment is not corrected.

Separate ignorance from incompetence. Not knowing something in a new field is a normal starting point, not a diagnosis. The problem is not the lack of knowledge, but the confidence that one knows when one does not. A productive question when entering a new field is: 'What exactly do I not know that I don't know?' followed by finding someone who knows the subject well enough to answer honestly.

Record what you learned, not just what you did. Imposter syndrome is fed by the fact that the person does not accumulate evidence of their competence per se: they notice the results (sales, clients, money), but they do not notice which specific decisions and skills led to them. Keeping records of which decision, based on which understanding, led to which result helps internalize the connection between competence and achievement.

Compare yourself with your past self, not with others. One of the mechanisms of imposter syndrome is comparing one's internal sensations with the external manifestations of other people. Others look confident because you see their exterior; you feel anxiety because you know your interior. This is a structurally dishonest comparison. Comparing yourself to your past self—where you were a year ago, two years ago—provides a more accurate signal of your actual trajectory.

What does not work

The idea of 'just be more confident' does not work. The problem is not the level of confidence per se, but its calibration. Unjustified confidence is as harmful as unjustified lack of confidence—just in different situations and with different consequences. The advice to 'believe in yourself' does not solve the problem of self-assessment accuracy.

The idea of 'read about imposter syndrome and remind yourself that you are not an imposter' does not work. The phenomenon is resistant to rational persuasion precisely because its nature is not rational—it is an attribution pattern (explaining the causes of success), not a logical error. It is treated not by knowing the concept, but by accumulating and internalizing evidence of competence.

Avoiding new areas to maintain accurate self-assessment does not work either. Metacognitive skills—the ability to accurately assess one's own competence—develop alongside subject competence, not instead of it. The only way to start seeing your blind spots in an area more accurately is to enter it, receive feedback, and adjust your self-assessment.

Conclusion

The Dunning-Kruger effect is real in its basic form—people inaccurately assess their competence, and low competence creates a higher risk of overestimation. The extreme version with the 'mountain of confidence' is an oversimplification, partially based on a statistical artifact. Imposter syndrome is real on the opposite end—and is found in highly effective people significantly more often than is commonly thought.

For an entrepreneur, both effects are equally important—and at different times. When entering a new area, there is the risk of overestimation and blind spots. As real achievements accumulate, there is the risk of underestimating them and avoiding the next step.

The general principle is the same: self-assessment requires an external anchor. It does not calibrate itself—it needs honest feedback, evidence recorded over time, and people who know the subject well enough to tell the truth.

---

Sources:

1. Kruger, J., & Dunning, D. (1999). Unskilled and Aware of It: How Difficulties in Recognizing One's Own Incompetence Lead to Inflated Self-Assessments. Journal of Personality and Social Psychology, 77(6), 1121–1134. https://doi.org/10.1037/0022-3514.77.6.1121

2. Nuhfer, E., Cogan, S., Fleisher, S., Gaze, E., & Wirth, K. (2017). How Random Noise and a Graphical Convention Subverted Behavioral Scientists' Explanations. Numeracy, 10(1). https://doi.org/10.5038/1936-4660.10.1.4

3. Gignac, G. E., & Zajenkowski, M. (2020). On the Relation Between Self-Assessed Intelligence and Measured IQ. Intelligence, 83, 101490. https://doi.org/10.1016/j.intell.2020.101490

4. Jansen, R. A., Rafferty, A. N., & Griffiths, T. L. (2021). A Rational Model of the Dunning–Kruger Effect Supports Insensitivity to Evidence in Low Performers. Nature Human Behaviour, 5(6), 756–763. https://doi.org/10.1038/s41562-021-01057-0

5. Clance, P. R., & Imes, S. A. (1978). The Impostor Phenomenon in High Achieving Women: Dynamics and Therapeutic Intervention. Psychotherapy: Theory, Research and Practice, 15(3), 241–247. https://doi.org/10.1037/h0086006

6. Bravata, D. M., Watts, S. A., Kiefer, A. L., et al. (2020). Prevalence, Predictors, and Treatment of Impostor Syndrome: A Systematic Review. Journal of General Internal Medicine, 35(4), 1252–1275. https://doi.org/10.1007/s11606-019-05364-1