The Dunning-Kruger effect, reported by David Dunning and Justin Kruger (1999), describes a metacognitive failure: people with poor skills in a domain tend to vastly overestimate their competence, while people with strong skills tend to slightly underestimate theirs. Those in the bottom quartile of performance typically rate themselves near the 60th-70th percentile. The irony is double: not only do they perform poorly, but their lack of skill robs them of the metacognitive ability to recognize their poor performance.
Mechanism
Dunning and Kruger argued that the skills needed to produce correct responses are the same skills needed to recognize what constitutes a correct response. A person who lacks grammatical knowledge cannot evaluate whether their writing is grammatically correct. A person who lacks logical reasoning skills cannot recognize when they have made a logical error. This dual burden — inability to produce and inability to recognize — creates a confident incompetence that is resistant to correction because the person cannot perceive the gap between their performance and competence.
The original Dunning-Kruger findings have been debated. Some researchers argue that the effect is partly statistical artifact: regression to the mean predicts that the lowest performers will overestimate (and the highest will underestimate) regardless of any metacognitive failure. However, Dunning and colleagues have provided additional evidence that metacognitive deficits are real: training the unskilled improves not only their performance but also their self-assessment, suggesting the original inability to self-evaluate was genuine rather than artifactual. The debate continues, but the core observation — that incompetence often comes with unwarranted confidence — is well-supported.