Children's selective trust: When a group majority is confronted with past accuracy.

In two experiments, 3- to 5-year-old children were tested for their preferences when seeking and accepting information about novel animals. In Experiment 1, children watched as two adults named unfamiliar animals - one adult was predominantly accurate, whereas the other was predominantly inaccurate, as judged by a teacher. In a subsequent test phase, participants viewed additional unfamiliar animals and were invited to endorse one of two conflicting names. Either the predominantly accurate or the predominantly inaccurate adult proposed one name, whereas a majority of three unfamiliar adults proposed the other name. Children were more likely to endorse the predominantly accurate adult as compared to the majority but showed no significant preference for the predominantly inaccurate adult as compared to the majority. In Experiment 2, participants watched two adults correctly name three familiar animals, but only one named three additional unfamiliar animals whereas the other expressed uncertainty. On subsequent test trials, children preferred the apparently well-informed adult to the less-informed adult but, contrary to the results of Experiment 1, children preferred the information provided by a majority instead of the apparently well-informed adult. The implications of these results are discussed in the light of previous research on children's selective trust in an accurate informant as compared to a consensus. Statement of contribution What is already known on the subject? Young children monitor past accuracy and use this epistemic cue to decide whom to trust; Children are receptive to information coming from a consensus; Non-epistemic cues, such as familiarity and accent, also influence children's deference What does this study adds? Children favour a dissenter over a majority if the dissenter's past accuracy has been publicly highlighted. They favour a majority if a dissenter's past accuracy has not been publicly highlighted. A confident informant is preferred to a hesitant informant.