Confidence is an asset—until it isn’t. It’s also a balancing act: Certainty is essential to moving forward, but too much of it can make it hard to see when we’re wrong. Add money to the mix, says a new study, and we run into trouble.
Accurately assessing what the study refers to as “the quality of our answers, actions or statements” is crucial for so many aspects of life. We want to know not only that our doctor or investment banker is confident about a recommendation but also that this confidence is itself reliable. That’s true for anyone we look to for advice, including ourselves.
But giving our decisions the exact level of confidence needed, as well as gauging its accuracy, is difficult. “We often overestimate the probability of being correct,” University of Amsterdam economist Maël Lebreton and his colleagues write in Science Advances.
And how does that estimation change when cash is riding on our choices? To find out, Lebreton and his colleagues enrolled 104 people in a series of visual tests using Gabor patches—black-and-white images that play tricks on the brain and are often used in experiments about perception. In each test, the participants had to decide which of two images had the higher contrast. Then, on a scale of 50 to 100 percent, they rated how confident they were in their answer. First, though, the researchers raised the stakes, telling the participants that they would win or lose money depending on the accuracy of their answers.
In part, the money helped. The thought of losing made some better at discriminating between correct and incorrect answers. But it also made participants who had rated themselves highly even surer of themselves. “The prospect of gains increased their overall confidence,” says Lebreton. “The prospect of losses decreased [it].” If a person already felt sure of their answer, the money didn’t add to that bias; it multiplied it.
Tomas Folke, a psychologist at the University of Cambridge, notes that because the incentives were fairly modest—ranging from 10 cents to $10—extrapolating more dramatic scenarios may be a stretch. It’s also unclear if the improved discrimination happened because the threat of loss sharpened people’s guesses or made them less confident.
Still, the findings raise questions about the best way to use monetary incentives. Offering experts large bonuses for an accurate decision, for example, “could paradoxically bias them into overconfident judgments,” says Lebreton. They also show how unreliable our daily judgment can be. “Numerous physiological and psychological factors bias us,” says Lebreton. Money, our surroundings and our upbringing are just a few. “We are far from being rational decision-makers.”