The Dunning-Kruger effect, first described in 1999 by David Dunning and Justin Kruger, suggests that people with low competence in a particular area tend to overestimate their ability. This effect has been widely popularized, often to explain why individuals may be unaware of their own shortcomings.

However, recent analysis suggests the celebrated effect might be an artifact of statistical interpretation rather than a genuine psychological bias. Dr. Ed Nuhfer and colleagues argued in 2016 and 2017 that the Dunning-Kruger effect could be replicated using random data. Their research indicated that both experts and novices underestimate and overestimate their skills with similar frequency, with experts doing so over a narrower range.

Further investigation by Dr. Patrick E. Mc Knight and Dr. Simone C. Mc Knight using statistical computing replicated these findings. They concluded that the effect, as originally observed, may not be a bias in human thinking but rather an artifact of how performance and self-assessment data are measured and visualized, particularly when separated into quartiles. The original experiment involved students assessing their performance on tests of grammar, humor, and logic, and comparing self-assessed scores to actual scores.