Bottom of the Class, Top in Their Own Minds
In 1999, psychologists David Dunning and Justin Kruger tested university students on logical reasoning, English grammar and even how well they could judge humour. Then they asked the students how well they thought they had done.
The weakest students were badly off. Their test scores placed them around the 12th percentile, meaning they did better than only about 12 percent of their peers. But on average, they ranked themselves in the 62nd percentile, comfortably above the middle.
This pattern is now called the Dunning–Kruger effect. It is usually described as a cognitive bias, a built-in habit of thinking that pushes our judgement wrong in a predictable way. It describes how people with low ability in a particular area tend to rate that ability far too highly.
Some researchers also use it for the flip side. People who are very good at something can underestimate their skill compared with others, possibly because they assume everyone else finds the task as easy as they do.
What It Does Not Mean
The effect is famous, and it is often misunderstood. In popular culture it gets boiled down to the idea that stupid people are too stupid to know they are stupid. That is not what the research says.
First, it is not about general intelligence. It is about skill at specific tasks, such as grammar, logic or driving, and says nothing about how clever someone is overall.
Second, it does not claim that unskilled people feel more confident than experts. Low performers do overestimate themselves, but their confidence is still lower than that of high performers.
The effect has since been found across a wide range of areas, including business, politics, medicine, driving, aviation, chess and school exams. Researchers usually measure it the same way. People complete a task, estimate how well they did, and their guess is compared with their real result. The gap tends to be biggest when people are asked to rank themselves against others, rather than simply guess their own score.
Why It Happens, Maybe
Dunning and Kruger's own explanation is known as the dual-burden account. Part of learning a skill is learning to tell good work from bad work. So people who lack a skill also lack the ability to see their own mistakes. They carry two burdens: they are not good at the task, and they cannot tell that they are not good.
Not everyone agrees. One major challenge says much of the effect may come from a statistical pattern called regression toward the mean. If you select the people with extremely low scores, their guesses about themselves will naturally sit closer to the middle. Add the better-than-average effect, our general habit of rating ourselves above average, and a similar pattern appears. On average, people think their IQ is 115, even though the average is 100.
Other explanations point to people's hopeful expectations going in, or to the fact that many low performers have very similar skill levels, which makes ranking yourself harder. Dunning has defended his work, arguing that the misjudgements are real whatever causes them.
When Confidence Costs You
The effect matters most when a wrong guess about yourself leads to a bad decision. Researchers have suggested it can push people toward careers they are not suited to. Overconfidence can also lead pilots to fly a new aircraft they are not properly trained for, or attempt moves beyond their ability.
It can cost money, too. A buyer who knows little about used cars may happily pay far too much, because they do not see the flaws and risks they are missing.
It is not all bad. Dunning has argued that overconfidence can help while you are carrying out a plan, by boosting motivation and energy. It hurts during planning, when you might ignore bad odds, take needless risks or skip preparing for problems.
The idea is older than the research. Charles Darwin wrote that ignorance more frequently begets confidence than does knowledge. And in 2000, Dunning and Kruger were awarded the Ig Nobel Prize, a joke award, for what the organisers called their modest report.
