More Than Just Agreement: Children’s Understanding That the Power of Consensus Stems From Group Composition

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

Abstract

We tend to rely on majority opinion (Corriveau & Harris, 2010; Mercier & Morin, 2019) but are often faced with contradictory perspectives, each shared by different consensus groups. Unless we have relevant expertise or access to information about how each group’s opinion was formed, we are left with only one source of evidence for determining which consensus group is more likely to provide valid information: the consensus members’ social identities.
Importantly, how social identity relates to the validity of a consensus depends on the type of information in question. When seeking to explain natural events, appealing to a diverse set of converging perspectives may be key; individuals’ experiences can serve as independent lines of evidence, and thus the conciliation of these experiences substantiates their shared view. In contrast, when seeking to explain a cultural phenomenon, a shared view among group members of that culture is likely more reliable since they have privileged access to information about that practice. To reliably and flexibly learn from consensus testimony, children need to assess how different social compositions of consensus groups affect reliability for learning about different types of information.
We present two experiments (Experiment 1: 72 5-6-year-olds; Experiment 2: 72 7-8-year-olds, 36 adults) investigating whether children use consensus members’ cultural group membership to select which consensus testimony to seek out and, if so, whether their use of this information varies as a function of the type of knowledge in question. Participants learned about a novel planet where various ‘weird’ things happen. Across 8 trials children heard about a ‘weird’ phenomenon, half naturally occurring (natural trial-type), and half cultural practices of a select alien group (cultural trial-type). On each trial, participants saw two alien groups – a diverse consensus group comprised of members of different cultures with different experiences or a homogenous group, comprised of members of the same culture – and asked to choose the best group to ask for an explanation. See Figure 1a for example stimuli for each trial type.
In Study 1, 5- and 6-year-olds reliably selected the homogenous group on cultural trials, but chose the diverse and homogeneous consensus groups equally on natural trials. There was no significant difference in performance across age groups, and no significant interaction between age group and trial type. In Study 2, 7-8-year-olds and adults aligned their consensus group choices with the type of phenomenon. On cultural trials, children and adults reliably selected the homogenous group. On natural trials, children selected the diverse consensus group more often than expected by chance, but less reliably than the adults. See Figure 1b for all results.
Taken together, this work suggests that shared cultural group membership among consensus members is a particularly powerful signal to credibility for questions relevant to that culture from early in development, but not until age 7-8 do children recognize the merits of diverse perspectives for learning about natural phenomena.

Authors

• Karen Carmel Levush, University of Maryland, College Park

• Jenna Alton, University of Maryland - College Park

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