After a brief look at the setup guide, he configured the smart speaker and announced it was done - music was playing in one room. He had no idea the other speakers in the network were silent, connected incorrectly due to a missed configuration step. From where he stood, the job was complete.
His sibling, who had installed home-theatre systems dozens of times, examined the same setup and felt uncertain - questioning whether the approach was correct, worrying that more complex integrations might expose gaps in their technique, despite a record of flawless previous installations.
One person had missed the error without noticing. The other had found no error but worried there was one.
What It Is
The Dunning-Kruger effect is a cognitive bias in which people with low ability at a task tend to overestimate their competence, while people with high ability tend to underestimate theirs.
Individuals scoring in the bottom quartile on assessments of specific skills - in the original research, tasks involving logical reasoning, grammar, and identifying humour - rated their own ability significantly higher than their actual performance. The overestimation is not random; it is a systematic feature of limited metacognitive skill.
The Mechanism
The effect arises from a deficit in metacognitive skill: those lacking competence also lack the ability to accurately assess their own performance. To recognise that you have made an error requires the ability to recognise what a correct answer looks like. When that ability is absent, the error is invisible, and the self-assessment cannot adjust.
Highly competent performers face a different problem. They are aware of the difficulty of the task. They are aware of the range of ways things can go wrong. This awareness can lead them to assume that others find the task similarly difficult - and to underestimate their standing relative to peers who are not holding the same level of complexity in mind.
What the Effect Actually Says
The Dunning-Kruger effect is widely cited and widely mischaracterised. Three clarifications are necessary.
It is a statistical tendency, not a universal rule. Not every person with low skill overestimates. Not every expert underestimates. The original research identified a pattern in group averages across controlled tasks - not a law of individual human behaviour.
It is not the same as impostor syndrome. The expert underestimation in the Dunning-Kruger literature is a relative effect: highly skilled performers tend to estimate their ability slightly above the group average but below their actual rank. This is not the same as impostor syndrome, which is a clinically recognised experience of persistent self-doubt and fear of being exposed as fraudulent.
It is not about intelligence in general. The effect has been demonstrated for specific skills - logical reasoning, grammatical knowledge, identifying humour in research tasks - not for general intelligence. Extrapolating to claim that it governs all domains and all kinds of competence goes beyond what the research shows.
Why It Matters
In self-directed learning. Early in acquiring a skill, the gap in metacognitive ability is largest - and therefore so is the potential for overconfidence. Feedback from objective external benchmarks, from peers, or from structured assessment provides information that self-assessment alone cannot.
In teams and organisations. People who are confident and articulate may be overestimating. People who are hesitant or self-deprecating may be the most capable. Neither confidence nor expressed uncertainty reliably signals actual competence.
In feedback design. If low performers cannot recognise their errors without being shown them, feedback processes that only confirm adequate performance will not improve metacognitive calibration. Showing the error, not just the correct answer, is more useful.