Testing What You’re Told: Young Children’s Empirical Investigation of a Surprising Claim

Abstract

Samuel Ronfarda* , Eva E. Chenb & Paul L. Harrisca University of Toronto, Mississauga, Canadab Hong Kong University of Science and Technology, Hong Kongc Harvard University, United StatesCONTACT Samuel Ronfard samuel.ronfard@utoronto.ca University of Toronto, Mississauga, 3359 Mississauga Road, CCT Building , Room 4059 Toronto, L5L 1C6, CanadaABSTRACTWe examined differences among children in their endorsement of an adult’s claim, their subsequent empirical investigation of that claim, and their resolution of any potential conflict between the claim and their empirical investigation. American and Chinese preschool (N = 171, M = 4.71 years) and elementary school (N = 128, M = 7.59 years) children were presented with five, different-sized, Russian dolls and asked to indicate the heaviest doll. Children typically selected the biggest doll. Children then heard either a false, counter-intuitive claim (i.e., smallest doll = heaviest) or a claim confirming their initial intuition (i.e., biggest doll = heaviest). Children frequently endorsed the experimenter’s claim even when it was counter-intuitive. The experimenter then left the room. During the experimenter’s absence, older children who had heard the counter-intuitive as opposed to the confirming claim explored the dolls more than younger children, especially when subtly prompted to explore. Moreover, only older children who heard the counter-intuitive claim simultaneously picked up the smallest and biggest doll, a more deliberate test of the experimenter’s claim. By implication, children engage in selective exploration following a surprising claim. Older children’s more systematic explorations of what they have been told may reflect improvements in their ability to test such claims and in their greater sensitivity to the fact that unexpected claims can and should be empirically investigated.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 40 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.