A Matter of Record: Children’s Evaluation of Knowledge Transmission for Text- and Audio-Based Knowledge Artifacts

• In Event: Reasoning About Epistemic Processes: Children’s Understanding of how Knowledge is Acquired, Transmitted and Socially Constructed

Abstract

As children develop, they gain increasing access to secondhand information sources, beginning with social others and expanding to a diverse array of knowledge artifacts, or nonhuman information sources such as books, audio recordings, and other informational media (e.g., Chandler-Campbell et al., 2022). Extensive research has established that young children assess verbal information by applying various social and epistemic criteria to its source (e.g., Tong et al., 2020), but less is known how they evaluate knowledge artifacts, which have limited information about their source, or author. There is some initial evidence that children may not understand the source of knowledge artifacts’ information, as many children privilege text-based information over verbal information in the absence of author information (e.g., Corriveau et al., 2014). In two experiments, we examine whether preschoolers (ages 4-6) use the knowledge or ignorance of an author to epistemically evaluate text-based and audio-based artifacts and whether they privilege artifact-based information in general for either artifact type.
For each experiment, children were shown video of a game where one of two balls (e.g., red or yellow) was hidden in a box. An author looked inside (knowledgeable author condition) or did not (ignorant author condition) and recorded “which ball [they thought was] in the box” via a written note (Experiment 1; n = 70) or an audio recorder (Experiment 2; n = 55) before leaving the room. Lastly, two informants entered the room, one of whom referenced the authored artifact (“This says it is the [red] ball”) and the other simply stated an answer (“I say it is the [yellow] ball”). Children were asked which ball was in the box and to justify their answer. After three trials, children were asked which informant provided better answers (informant preference) and their assessment of author knowledge to examine whether these predicted their likelihood to endorse the artifact’s answer in a given trial.
In both experiments, regardless of age, children used two independent evaluations when determining their endorsements, privileging knowledge artifacts in general but also considering author knowledge. In Experiment 1, children were 3.07 times more likely to endorse the text artifact answer when they indicated a text informant preference (β = 1.12, SE = 0.35, p < 0.001) and 2.10 times more likely in the knowledgeable author condition (β = 0.74, SE = 0.33, p = 0.03). In Experiment 2, children were 8.79 times more likely to endorse the audio artifact answer when they indicated an audio informant preference (β = 2.17, SE = 0.66, p < 0.001) and 2.85 times more likely in the knowledgeable author condition (β = 1.05, SE = 0.38, p = 0.006). Coding is underway to examine whether children’s endorsement justifications indicate whether they considered certain elements of the knowledge transmission or construction process in their reasoning.
Overall, results indicate preschoolers consider author knowledge relevant to the reliability of knowledge artifacts’ information but display a stronger tendency to prefer both types of knowledge artifact regardless of author knowledge.

Authors

• Ian L. Chandler-Campbell, University of Texas at Dallas
  Presenting Author

• Maliki E. Ghossainy, Boston University

• Kathleen H. Corriveau, Boston University