Confidence may be a 'measurable quantity': study

Confidence may actually be a measurable activity in the brain, not just an emotion or a feeling, scientists have found in experiments with rats.
Researchers at Cold Spring Harbor Laboratory (CSHL) in the US have identified a brain region in rats whose function is required for the animals to express confidence in their decisions.
The team, led by CSHL Associate Professor Adam Kepecs, devised a method to study decision making in rats.
The rats were offered an odour that they were trained to associate with one of two doors. When they chose the correct door, they were rewarded. This part was easy for the animals: their selections were almost always correct.

Kepecs and his team then offered a mixture of the two scents, with one dominating over the other by only a very small percentage.
The rats now needed to choose the door representing the dominant odour in order to get their reward - a choice that reflects their best guess.

Researchers said confidence can be measured simply by challenging a rat to wait for the reward to be revealed behind the door. The time they are willing to wait serves as a measure of the confidence in their original decision.
"We found that the rats are willing to 'gamble' with their time," Kepecs said, sometimes waiting as much as 15 seconds, which is an eternity for these animals.

"This is something that we can measure and create mathematical models to explain. The time rats are willing to wait predicts the likelihood of correct decisions and provides an objective measure to track the feeling of confidence," said Kepecs.
The researchers hypothesised that a distinct region of the brain might control confidence.
Previous work has suggested that the orbitofrontal cortex (OFC), a part of the brain involved in making predictions, might have a role in decision confidence.
Kepecs and his team specifically shut off neurons in the OFC, inactivating it, and found that rats no longer exhibited appropriate levels of confidence in their decisions.

"With an inactive OFC, the rats retained the ability to make decisions - their accuracy did not change," said Kepecs.
"And they spent the same amount of time waiting for a reward on average. The only difference is that animals' willingness to wait for a reward was no longer guided by confidence. They would often wait a long time even when they were wrong," he said.
"We now know that the OFC is critical for making on-the-fly predictions in rats. The human OFC is just a more sophisticated version of the rodent counterpart," Kepecs said.
The team is expanding their research to explore how the elusive feelings of confidence are based on objective predictions that influence human decisions as well.

The study was published in the journal Neuron.