Knowing to Ask and Asking to Know: Lessons from Children’s Question-Asking and Selective Trust

• In Event: Poster Session 06
  In Poster Session: PS 06 Topic - Social Cognition

Abstract

Children ask questions constantly, but their skills in effective, strategic questioning do not emerge until later in their school years. However, prior work shows that children can identify more efficient questioning strategies in others, even before they can generate their own, and can use this information to guide their learning (e.g., Herwig, 1982; Ronfard et al., 2018; Rothe et al., 2018; Ruggeri et al., 2017). The present study sought to (1) replicate the finding that young children can distinguish between others’ more and less effective questioning strategies; and (2) test whether children use this distinction to determine whether efficient questioners are also more reliable, more knowledgeable, and broadly more competent.
In this online study, children (N = 160, 4-7-year-olds) participate in a four-trial video-based Question Game and are tasked with identifying which one of two puppet characters is the more efficient, capable questioner. In each trial, the two puppet characters direct their questions to a third puppet: one always asking broad questions to limit the scope of possible answers (more efficient), and the other always asking narrowly-focused questions that can only eliminate one possible answer at a time (less efficient) (see Figure 1). Following the Question Game, children are asked to identify which of the questioners is most likely to know how a new toy works (reliability trial), be a better teacher (knowledgeability trial), and help them fix a broken toy (competency trial) (see Figure 2). Across all trials, children are asked to justify their selections.
There are several planned analyses:
For the Question Game: One-sample t-tests will be used to determine whether choice of competent questioner (i.e., the broader, more efficient questioner) occurs more often than chance; One-way ANOVAs will be used to determine whether this proclivity differs across age groups.
For the latter three trials: Binomial tests will be used to determine whether children indicated the more competent questioner as also more reliable, knowledgeable, and more competent; Logistic regressions will be fit to determine whether Question Game performance predicts children’s later selections.
For the justifications: Justifications will be coded with respect to children’s explicit reference to the quality of the characters’ questioning strategies; ANOVAs will be used to determine differences by age group, ordinal logistic regressions will be used to determine whether justification quality predicts children’s abilities to identify the more competent questioner; Logistic regressions will be fit to determine whether justification quality predicts their choice of a more competent questioner across each test trial.
Pilot data suggests a promising trend, confirming (1) children’s successful identification of the more-competent questioner, and (2) children’s generalization of this attribute across other cognitive capabilities, such as knowledgeability and broader competency. The study is expected to complete data collection by February of 2021. This work has implications for our understanding of how children search for information, and the assessments they make of the informants who provide it; In turn, this has consequences for the development of later, higher-order science and information literacy.

Authors

• Hailey Gibbs, University of Maryland - College Park
  Presenting Author

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