Poster #198 - Stability of Individual Differences in Preschoolers’ Endorsement of Accurate Informants Across Selective Social Learning Tasks

• In Event: 3-053 - Poster Session 10
  In Poster Session: PS 10 Section - Social Cognition

Integrative Statement

Background: Social learning serves as the foundation of a child’s understanding of the world around them. Learning from different sources of information is vital; however, not all informants are equally trustworthy. The ability to discern reliable sources from others, and proceed to selectively learn from the reputable source is thus an essential skill in development. Children have repeatedly been shown to selectively learn from previously accurate informants (see Harris, Koenig, Corriveau & Jaswal, 2018 for a detailed review), but the stability of individual differences has yet to be studied.

Objectives: The main goal of this study is therefore to assess whether score differences found between individual children on any given selective learning task reflect stable individual differences. In other words, this research explores whether individual scores across selective learning tasks are related to different ability levels, or if on the contrary individual performance is highly variable across situations and methodological variations.

Methods: To asses the stability of individual differences, 3- to 5-year-old children (n=102) completed selective learning tasks based on prior accuracy at two time points: either identical selective learning tasks conducted one week apart (Study 1) or parallel selective learning tasks conducted one immediately after the other (Study 2). All selective learning tasks consist of a familiarization and an assessment phase. During the familiarization phase two informants name four common objects (e.g. crayon), in which one informant is consistently accurate and the other inaccurate. During the assessment phase both informants name four novel objects with contradictory nonsense labels (e.g. “Misk” versus “Jole”) and the child is asked what they think each novel object is called. One point was awarded each time the child endorsed the label provided by the previously accurate informant.

Results: In each study, one-sample t-tests were conducted in order to compare the mean total score of each selective learning task to chance. Results suggested that, as expected and replicating prior research, children sided significantly with the previously accurate informant in both studies and on both administrations (means between 2.32 and 2.69 out of 4, ps <.03). To test for the stability of individual differences, correlation coefficients were calculated: In Study 1, performance on the first and second administrations were not correlated (r= .14, p= .166, one-tailed, ns), in Study 2 performance on parallel versions were significant but with a very small correlation (r= .23, p= .048, one-tailed). This suggests that individual stability is at best low and at worst absent.

Conclusion: In summary, children tend to preferentially rely on a previously accurate informant over an inaccurate one when learning new information. However, individual scores on these tasks are not strongly consistent over time or across parallel versions of the tasks, indicating that selective learning is highly situation-dependent. Thus, it is of utmost importance that researchers take precaution when interpreting individual differences in performance on selective learning tasks.

Authors

• Sophie Fobert, University of Ottawa
  Presenting Author

• Isabelle Cossette, University of Ottawa
  Non-Presenting Author

• Patricia Brosseau-Liard, University of Ottawa
  Non-Presenting Author