David Dunning

Introduction

David Alan Dunning (born 1960) is an American social psychologist, a professor of psychology at the University of Michigan and a retired professor at Cornell University. He is known above all for the paper he wrote with his graduate student Justin Kruger in 1999, "Unskilled and Unaware of It," which reported that the people who perform worst on a task are often the most confident about how well they did. The finding, now called the Dunning-Kruger effect, has become one of the most widely cited ideas in the psychology of judgment and one of the few research results to enter everyday speech. Dunning has spent much of his career on the broader question behind it: why people's pictures of themselves diverge so persistently from the evidence, and what that costs them. The effect has also become a focus of methodological argument, with critics contending that the classic evidence is largely a statistical artifact and Dunning and his collaborators responding with new tests and a large-scale review.

Early Life and Education

Dunning took his bachelor's degree in psychology at Michigan State University in 1982 and his doctorate at Stanford University in 1986, where he worked with Lee Ross, a central figure in the study of attribution and social judgment. His dissertation, "Situational construal and sources of social judgment," addressed a question that recurs throughout his later work: how people interpret the situations they are in, and how those interpretations drive the conclusions they reach about themselves and others.

Academic and Professional Career

Dunning spent most of his career at Cornell University before moving to the University of Michigan, where he is a professor of psychology and a faculty associate at the Research Center for Group Dynamics in the Institute for Social Research. He has served as executive officer of both the Society for Personality and Social Psychology and the Foundation for Personality and Social Psychology, and as an associate editor of the Journal of Personality and Social Psychology. In 2021 Stanford University's standardized citation database listed him among the most-cited psychological scientists in the world, and in 2023 he and Kruger shared the University of Louisville Grawemeyer Award in Psychology for the idea behind the Dunning-Kruger effect.

His research program has consistently returned to self-assessment: how people estimate their own performance, why those estimates are biased, and what happens when the bias is corrected. He has applied the question to emotional intelligence in managers, to the confidence of beginning learners, to the gap between what people predict they will do and what they actually do, and to the way chronic self-views shape estimates of performance independently of any fact about it.

Major Contributions to Psychology

The Dunning-Kruger effect

The 1999 paper reported four studies in which participants completed tests of humor, logical reasoning, and English grammar and then estimated how well they had done relative to the rest of the group. Participants in the bottom quartile performed at roughly the twelfth percentile, but placed themselves around the sixty-second: they believed they had scored above average. Participants in the top quartile were more accurate about their absolute performance but underestimated their standing relative to others, apparently assuming that tasks they found easy were easy for everyone. The authors' explanation was that the skills needed to produce a correct answer are, to a substantial degree, the same skills needed to recognize a correct answer, so incompetence carries a double burden: it produces errors and simultaneously removes the capacity to detect them. In a fifth study, training participants in the task improved both their performance and the accuracy of their self-assessments, supporting the causal claim.

The idea came from a news story about a man who robbed banks with lemon juice smeared on his face, having reasoned that lemon juice, used as invisible ink, would make him invisible to surveillance cameras. Dunning has said he thought the paper would never be published and has expressed surprise at how thoroughly the idea escaped the laboratory.

Why the unskilled are unaware

Subsequent work tested and extended the mechanism. Studies with Joyce Ehrlinger, Kerry Johnson, and Justin Kruger showed that poor performers remain overconfident in realistic settings and even when offered money for accuracy, and that the effect is not merely a matter of motivation. Other work showed that people's global beliefs about their own ability can drive performance estimates independently of actual performance, and that manipulating which self-view seems relevant changes estimates without changing achievement. With Oliver Sheldon and Daniel Ames, Dunning found that the same overoptimism extends to self-assessments of emotional intelligence, a domain in which feedback is frequent but unusually ambiguous.

Flawed self-assessment in practice

With Chip Heath and Jerry Suls, Dunning reviewed the consequences of miscalibration for health, education, and the workplace, arguing that the costs fall unevenly: the people least able to judge their own competence are often the ones making consequential decisions — a patient deciding whether to seek a second opinion, a student deciding whether to study, an official deciding whether to trust a forecast. The practical implication drawn from this line of work is that calibration is not a talent but a product of feedback environments, and that making standards external, observable, and repeated improves it.

Major Publications

Self-Insight: Roadblocks and Detours on the Path to Knowing Thyself (2005)
Kruger, J., & Dunning, D. (1999). Unskilled and unaware of it: How difficulties in recognizing one's own incompetence lead to inflated self-assessments. Journal of Personality and Social Psychology, 77, 1121-1134.
Dunning, D., Johnson, K., Ehrlinger, J., & Kruger, J. (2003). Why people fail to recognize their own incompetence. Current Directions in Psychological Science, 12, 83-87.
Dunning, D., Heath, C., & Suls, J. M. (2004). Flawed self-assessment: Implications for health, education, and the workplace. Psychological Science in the Public Interest, 5, 69-106.
Ehrlinger, J., Johnson, K., Banner, M., Dunning, D., & Kruger, J. (2008). Why the unskilled are unaware? Further explorations of (absent) self-insight among the incompetent. Organizational Behavior and Human Decision Processes, 105, 98-121.
Dunning, D. (2011). The Dunning-Kruger effect: On being ignorant of one's own ignorance. In Advances in Experimental Social Psychology (Vol. 44, pp. 247-296).
Schlösser, T., Dunning, D., Johnson, K., & Kruger, J. (2013). How unaware are the unskilled? Empirical tests of the "signal extraction" counter-explanation for the Dunning-Kruger effect in self-evaluation of performance. Journal of Economic Psychology, 39, 85-100.

Impact and Legacy

Few psychological findings have travelled as far. The Dunning-Kruger effect is cited in discussions of education, medicine, management, and politics, and the 1999 paper has been referenced in thousands of scholarly works. The Grawemeyer committee, awarding the 2023 prize, argued that the idea matters most in precisely the circumstances that produced it: an information environment in which people must decide when to trust experts and when to trust themselves. Dunning's broader contribution was to make self-assessment accuracy a research topic in its own right, with identified mechanisms and measurable consequences, rather than a curiosity.

The empirical status of the effect is genuinely disputed. Critics have argued for two decades that the standard method — plotting self-estimates and objective scores across quartiles — is confounded by two well-known phenomena: the better-than-average effect, since almost everyone places themselves above the midpoint regardless of ability, and regression toward the mean, since measurement error alone pushes extreme scores and extreme self-estimates in opposite directions. Joachim Krueger and Richard Mueller showed in 2002 that these two ingredients can generate the pattern without any special metacognitive failure, and Marian Krajc and Andreas Ortmann offered a related signal-extraction account. The most direct challenge came in 2020, when Gilles Gignac and Marcin Zajenkowski argued that the conventional test is confounded and applied two inferential alternatives — a test for heteroscedasticity and nonlinear regression — to a large community sample who rated their own intelligence and completed the Advanced Raven's Progressive Matrices. They found no significant heteroscedasticity and an essentially linear relationship, and concluded that while the phenomenon may be real for some skills, its magnitude is much smaller than reported. A follow-up analysis by Gignac in 2024 put the effect's influence on a limited segment of the population at negligible levels.

The artifact case is not the last word. Working with a British birth-cohort study, Anna Adamecz, Radina Ilieva and Nikki Shure (Journal of Behavioral and Experimental Economics, 2025) constructed composite objective-ability scores from eighteen tests administered at ages 5, 10 and 16 and compared them with self-estimates of school performance and of being clever. Replicating the models of Gignac and Zajenkowski (2020) and of Dunkel, Nedelec and van der Linden (2023) on far better-measured data, they report that the effect survives — which matters because measurement of ability is precisely where the earlier critiques located the weakness, and this is the first test using composite rather than single-occasion measures. They also report what they describe as the first test of heterogeneity by gender: the Dunning-Kruger effect itself turned out not to be gender-specific, but the self-assessment bias behind it is, arising in their data from men overestimating and women underestimating their relative standing. A related objection, advanced by Robert McIntosh and Sergio Della Sala, holds that the pattern is better described as noise plus a general better-than-average bias than as a distinct metacognitive failure; that account has not been settled either.

Dunning and his collaborators have replied rather than ignored the critique. His 2011 review in Advances in Experimental Social Psychology explicitly addressed the theoretical controversies over interpretation and distinguished the errors of poor performers from those of top performers. With Thomas Schlösser and colleagues he ran direct empirical tests of the signal-extraction counter-explanation — the claim that poor performers are simply noisier, not metacognitively impaired — and reported results supporting the original account, including evidence that poor performers are not merely guessing but are systematically misreading their own performance. Dunning has also distanced himself from the popular use of the term as an insult, noting that the effect is a statistical tendency within groups and not a diagnosis of any individual, and that the same research shows experts erring in the opposite direction. The result is an active, unresolved methodological argument in which both sides now agree on at least one point: the magnitude of the effect has been overstated in the telling, and the question is what, if anything, remains once the artifacts are removed.