Poster #200 - You Sound Like you Know, but did you Check? Children’s Understanding That Generic Claims Require Sufficient Verification

• In Event: 3-003 - Poster Session 09
  In Poster Session: PS 09 Section - Social Cognition

Integrative Statement

Children are prodigious social learners, and are equipped with a repertoire of skills that enable them to reason about information and informants (see Mills, 2013 for a comprehensive review). Previous work examining children’s evaluations of others’ claims found that young children prefer informants who fully verify their claims over those who do not, or who do so insufficiently (Butler, Schmidt, Tavassolie, & Gibbs, in press). However, previous literature has also demonstrated powerful effects of generic language, language that refers to whole kinds or categories, on the inductive inferences both children and adults will make about kinds and about available evidence, and which has been shown to require less evidence to be taken as true (Cimpian, 2012), and have marked implications for people’s predictions about new members of a kind (Cimpian & Markman, 2011).
The current study examines whether, when faced with claims made using generic language, children are able to apply the same empirical reasoning process in evaluating others’ inductive inferences.
Ninety-six 4- to 6-year-olds were shown digital puppet shows, showing puppets investigating whether a set of novel objects, “blickets,” made a novel toy play music. One puppet checked all four of four available blickets, while the other checked only one of the four blickets, but did so four times. Both puppets then made the same generic claim, “blickets make the machine play music.” Thus, both puppets had generated equal amounts of evidence, but only one generated evidence that might be considered sufficient to make this generic claim. Children were then asked to assess the validity of the puppets’ rational inference regarding the function of the toy and to provide justifications for their responses. Following an additional puppet show, structurally identical to the first, children were asked to consider the claim itself and to provide additional justifications.
Logistic regressions within trial type with age as a predictor and choice of informant (verified or insufficiently verified) as the outcome were conducted. The rational inference trial was nonsignificant (X2 (63) = 40.859, p= 0.484), as was the generic claim trial (X 2 (63)= 42.659, p= 0.099), indicating that age is not a significant predictor of children’s selectivity on the basis of verification behaviors. Thus, the age groups were collapsed and considered together in a follow-up binomial test for each trial type. Findings indicate that children across all ages were significantly more likely than expected by chance to select the verifying puppet across both the rational inference (p=.001) and generic claims (p<.001) trials. Indeed, roughly ⅔ (65.98%) of all children (ages 4- through- 6-years) selected the verifying puppet on both trial types.
Children are heavily dependent on testimony in building a reliable knowledge base, and must thus practice epistemic vigilance, irrespective of how the claims they encounter are communicated. These findings indicate that young children can effectively use their developing understanding of how inductive inferences are generated in order to reason about the validity of others’ inferences, even when the claims they make about those inferences are communicated generically.

Authors

• Hailey Gibbs, University of Maryland College Park
  Presenting Author

• Lucas Payne Butler, University of Maryland, College Park
  Non-Presenting Author