What Is the Dunning–Kruger Effect?
The IQ Town Team
Psychometrics & test design
The Dunning–Kruger effect, as the 1999 paper framed it, is that people who do badly on a test often think they did fine, because the skill you need to perform is close to the skill you need to judge the performance. The popular version — fools are certain, experts are humble — is a cartoon of that result.
Kruger and Dunning, 1999
- Tests of humor, grammar, and logic
- Bottom quartile wildly overestimated their percentile
- Top quartile slightly underestimated
- A training session improved the poor judges
What traveled online
- 'Idiots think they are brilliant'
- A mountain chart that is not in the paper
- Used as an insult, not a measurement
- Ignores later statistical critiques
What the 1999 paper found
David Dunning and Justin Kruger had people take short tests and then estimate how they did relative to others. The lowest performers placed themselves near the middle. They were not claiming to be world-class. They were failing to see that they had failed. The authors argued that the missing skill was the same skill the test required: if you cannot see a grammatical error, you also cannot see that you make them.
What the meme added
A chart of confidence soaring while competence is low, then a valley of despair, then expertise. That chart is pedagogy from later talks, not the study's figure. Later researchers have shown that a chunk of the 'effect' appears even in simulations where people guess with noise, because a low true score has more room to be overestimated and a high one has more room to be underestimated. The debate is about how much is psychology and how much is arithmetic. It is not a license to diagnose strangers.
This is cousin to why smart people believe false things: accuracy and confidence are not the same dial. Critical thinking is the skill of checking either one.
Guess how sure you are — then check
Try one example before you start the full test. Instant feedback — no account required.
All blips are blops. Some blops are blups. Which conclusion must be true?
A score beats a self-assessment
The free test will not ask you how smart you feel. It will score the items.
How to use the idea
Get an external score when the skill matters. The margin of error means even the external score is a band, which is still better than a vibe. If you train the actual skill, your ability to judge it tends to improve too — that was part of the original paper, and it is the useful part. Take the free IQ test if the hunch you want to check is about reasoning. Then believe the band more than the feeling you had before you started.
Aim the effect at yourself
The finding is a reason to seek feedback. It is a weak insult. You cannot see someone else's metacognition from a comment thread.Frequently asked questions
What is the Dunning–Kruger effect?
It is the pattern, reported in 1999, that low performers on a test overestimate how well they did, because judging the performance uses the same skill as producing it. The pop version overstates the certainty and the drama.
Do experts think they are incompetent?
In the original studies, high performers slightly underestimated their percentile. They did not believe they were incompetent.
Is Dunning–Kruger just a statistics artifact?
Part of the classic graph can be produced by noisy guesses plus real scores. That critique shrinks the effect. It does not make self-assessment reliable.
Replace the hunch
About 10 minutes, unassisted.