Dubbed as the theory of mind, the ability to infer another’s perspective – emotional, intellectual, or visual – boosts with age.

To reach the conclusion, the researchers made kids watch two puppets – Sally and Anne – play with a marble, then put the marble back in a box.

Anne "left" and Sally grabbed the marble, played with it, and then returned the marble instead to a bag.

Where will Anne first search for the marble, the researchers asked the children as part of the study.

"Before four, kids say she’s going to look in the bag, but after four they know she has a false belief," New Scientist quoted Iroise Dumontheil, a cognitive neuroscientist at University College London, UK, who led the new study.

However, Dumontheil said, brain scans suggest that a teenage mind toils harder when inferring the outlook of others, compared with adults.

And a brain region implicated in theory of mind, the medial prefrontal cortex, continues to develop through adolescence, the scientist added.

To see if there is a behavioural consequence of these biological changes, she and colleagues tested children, adolescents, and adults on their ability to infer the spatial perspective of another person in a simple computer game.

Volunteers - 179 females ranging in age from 7 to 27 - saw a bookshelf with a variety of different sized balls and other objects on four different rows.

Few of the objects sit in front of opaque backgrounds, obscured to someone standing on the other side of the shelf, while some sit in front of a see-through background.

Participants were asked to adopt the perspective of a man standing on the other side of the shelf and move the small ball to the left, using a mouse. In a typical test, a golf ball and tennis ball are both visible to the participant, but the golf ball is obscured from the point of view of the observer.

The correct response, then, is to move the tennis ball.

Kids under the age of 10 moved the wrong ball in about three-quarters of trials. Children aged 10 through 13 scored marginally better, and teens answered wrong on two-thirds of trials. Adults, however, did better than 50-50, on average.