Over the past years there has been a growing interest in children gullibility or absence thereof. More and more people try to understand the way children calibrate their trust, and ask at what age they start to evaluate the quality of different means of gaining knowledge (perception, inference, communication) or the quality of different speakers (benevolence, competence) and to base their choices on these evaluations (see the work by Paul Harris and Fabrice Clément for example).

In this line of study, we can ask if children rely on a general heuristic that adults should be trusted more than children or if they are able to tailor their choices to different factors, for example the competence of the characters. A new study has pitted these two possibilities in the following way: children (3 and 4-year-olds) were shown movies in which two characters, a children and an adult were naming known objects. Sometimes they were both accurate, sometimes only one of them was, and some other times both were inaccurate. The last action involved a new object, unknown to the child, that was named in different ways by the two characters. The question then is: what name will the child choose? When both speakers were accurate, they mostly choose the name given by the adult (quite a sensible thing to do). However, when only one of the speakers was accurate, they choose the name given by the accurate speaker even if it was the child. They were thus able to reverse their preference because the adult as proven to be unreliable in a few naming situations.

So adults, beware: if you make too many mistakes, children will stop listening to you!