How Children Change Their Beliefs in Light of Disagreement

• In Event: Mechanisms underlying imitation, social learning, and social preferences

Abstract

A healthy public discourse requires that people respond to disagreement in reasonable ways. For example, a reasonable person should update their belief after learning that others have better evidence for an alternative view. Here, we investigate children’s developing ability to selectively change or maintain their initial belief when confronted with a disagreeing peer.

With age, children are increasingly able to use different epistemic cues when forming beliefs: they assess how many informants support a claim (Corriveau, Fusaro, & Harris, 2009), and monitor the reasons people offer in support of their beliefs (e.g. Koenig, 2012), to name just two examples. In situations of disagreement, however, agents have typically formed a prior belief and must respond to an alternative belief. According to Frances (2014), a reasonable way to respond is as follows: if you have good reason to think you have an epistemic advantage over the person you are disagreeing with, you should “stick to your guns”. If the other person has an epistemic advantage, you should adopt their view; and if the two of you are epistemically roughly equal, you should suspend judgment.

We tested whether participants in three age groups (children aged 4-6, 7-9, and adults; N=114) adhere to these principles in their responses to disagreement, using an online picture book paradigm. In each of three conditions, participants formed an initial belief about where a pet bunny went and were then exposed to a disagreeing peer, who had developed a belief that opposed the participant’s belief. Between conditions, we varied whether the evidence supporting the participant’s belief was better than (condition 1), equal to (condition 2), or worse than (condition 3) that of the other person. Participants then had the option of either sticking with their initial belief, suspending judgment, or adopting the other’s belief.

Overall, we found that participants responded reasonably to the disagreements. Specifically, all age groups were more likely to stick with their initial belief when their evidence was better than the other’s evidence, and to adopt the other’s belief when their own evidence was worse (p’s<.01; Figure 1). Thus, some ability to respond reasonably to disagreements is present from at least the preschool years. However, we also found two interesting developmental differences: first, overall, younger children adopted the other person’s belief significantly more often than older children (p=.02). Second, unlike older children and adults, 4- to 6-year-olds did not yet demonstrate the “intellectually virtuous” (Haney, 1964) ability to suspend judgment in the equal evidence condition (Figure 2).

Possible explanations for these developmental differences will be discussed. One reason for younger children’s difficulty with suspending judgment might be that our operationalization of this response option was too complex and abstract. Previous work from the meta-cognition literature suggests that while young children sometimes have difficulties making explicit meta-cognitive judgments, they show evidence of implicitly monitoring their uncertainty in their information seeking behavior (e.g. Hembacher, deMayo, & Frank, 2017). Our planned work explores whether young children may show evidence of suspending judgment in their search for additional evidence.

Authors

• Antonia Frederike Langenhoff, University of California - Berkeley
  Presenting Author

• Mahesh Srinivasan, University of California - Berkeley
  Non-Presenting Author

• Jan Engelmann, University of California - Berkeley
  Non-Presenting Author