Incompetent People Most Likely To Be Confident Study Shows

March 28th, 2006

There are many incompetent people in the world. Dr. David A. Dunning is haunted by the fear that he might be one of them.

Dunning, a professor of psychology at Cornell, worries about this because, according to his research, most incompetent people do not know that they are incompetent.

On the contrary. People who do things badly, Dunning has found in studies conducted with a graduate student, Justin Kruger, are usually supremely confident of their abilities—more confident, in fact, than people who do things well.

``I began to think that there were probably lots of things that I was bad at, and I didn’t know it,’’ Dunning said.

One reason that the ignorant also tend to be the blissfully self-assured, the researchers believe, is that the skills required for competence often are the same skills necessary to recognize competence.

The incompetent, therefore, suffer doubly, they suggested in a paper appearing in the December issue of the Journal of Personality and Social Psychology.

SanFranciscoGate

``Not only do they reach erroneous conclusions and make unfortunate choices, but their incompetence robs them of the ability to realize it,’’ wrote Kruger, now an assistant professor at the University of Illinois, and Dunning.

This deficiency in ``self-monitoring skills,’’ the researchers said, helps explain the tendency of the humor-impaired to persist in telling jokes that are not funny, of day traders to repeatedly jump into the market—and repeatedly lose out—and of the politically clueless to continue holding forth at dinner parties on the fine points of campaign strategy.

In a series of studies, Kruger and Dunning tested their theory of incompetence. They found that subjects who scored in the lowest quartile on tests of logic, English grammar and humor were also the most likely to ``grossly overestimate’’ how well they had performed.

In all three tests, subjects’ ratings of their ability were positively linked to their actual scores. But the lowest-ranked participants showed much greater distortions in their self-estimates.

Asked to evaluate their performance on the test of logical reasoning, for example, subjects who scored only in the 12th percentile guessed that they had scored in the 62nd percentile, and deemed their overall skill at logical reasoning to be at the 68th percentile.

Similarly, subjects who scored at the 10th percentile on the grammar test ranked themselves at the 67th percentile in the ability to ``identify grammatically correct standard English,’’ and estimated their test scores to be at the 61st percentile.

On the humor test, in which participants were asked to rate jokes according to their funniness (subjects’ ratings were matched against those of an ``expert’’ panel of professional comedians), low-scoring subjects were also more apt to have an inflated perception of their skill. But because humor is idiosyncratically defined, the researchers said, the results were less conclusive.

Unlike unskilled counterparts, the most able subjects in the study, Kruger and Dunning found, were likely to underestimate their competence. The researchers attributed this to the fact that, in the absence of information about how others were doing, highly competent subjects assumed that others were performing as well as they were—a phenomenon psychologists term the ``false consensus effect.’’

When high-scoring subjects were asked to ``grade’’ the grammar tests of their peers, however, they quickly revised their evaluations of their own performance. In contrast, the self-assessments of those who scored badly themselves were unaffected by the experience of grading others; some subjects even further inflated their estimates of their own abilities.

``Incompetent individuals were less able to recognize competence in others,’’ the researchers concluded.

In a final experiment, Dunning and Kruger set out to discover if training would help modify the exaggerated self-perceptions of incapable subjects. In fact, a short training session in logical reasoning did improve the ability of low-scoring subjects to assess their performance realistically, they found.

The findings, the psychologists said, support Thomas Jefferson’s assertion that ``he who knows best knows how little he knows.’’

And the research meshes neatly with other work indicating that overconfidence is common; studies have found, for example, that the vast majority of people rate themselves as ``above average’’ on a wide array of abilities—though such an abundance of talent would be impossible in statistical terms. This overestimation, studies indicate, is more likely for tasks that are difficult than for those that are easy.

Such studies are not without critics. Dr. David C. Funder, a psychology professor at the University of California at Riverside, for example, said he suspects that most lay people have only a vague idea of the meaning of ``average’’ in statistical terms.

``I’m not sure the average person thinks of `average’ or `percentile’ in quite that literal a sense,’’ Funder said, ``so `above average’ might mean to them `pretty good,’ or `OK,’ or `doing all right.’ And if, in fact, people mean something subjective when they use the word, then it’s really hard to evaluate whether they’re right or wrong, using the statistical criterion.’’

But Dunning said his current research and past studies indicated there are many reasons why people would tend to overestimate their competency and not be aware of it.

In various situations, feedback is absent, or at least ambiguous; even a humorless joke, for example, is likely to be met with polite laughter. And faced with incompetence, social norms prevent most people from blurting out ``You stink!’’—truthful though this assessment may be.