Young Children Infer Relative Competence Based on Efficiency of Process and Quality of Outcome

• In Event: The Development of Children’s Understanding of Competence and Achievement

Abstract

Beliefs about our own and others’ competence guide everyday decisions. To decide whom to ask for help or how best to collaborate, children must reason about other people’s abilities and how good they are relative to others. However, competence is an abstract concept that is not directly observable; instead, it can be gleaned from the objective properties of the task completed by an agent (i.e. task-based features, like task difficulty) as well as the subjective properties of the agent (i.e. agent-based features, like speed). Prior work, however, has often used explicit verbal cues to describe the relevant features or experimental tasks that confounded these features, leaving open questions about how children acquire an integrated notion of competence. Here, we examine how 4- to 6-year-olds evaluate the relative competence of two agents by using a novel paradigm that allows us to systematically vary task-based and agent-based features without using explicit linguistic or gestural cues.

Across 5 Experiments, children watched videos of two agents who built block towers. We manipulated task difficulty (e.g., 10-block vertical tower vs. 10-block horizontal line, see Figure 1) and agents’ time-to-completion (e.g., one finishes building before the other). Critically, we never provided verbal cues to indicate that one was “better” or “faster” than the other, and masked the building process to eliminate direct perceptual cues to agents’ relative speed. The final building outcomes were revealed after both agents indicated they were done building by raising their hands and moving to the side of the occluder.

When task-based difficulty was matched (agents built the same tower), 4-5-year-olds successfully used agent-based differences in completion speed to infer relative competence (Exp.1: n = 30; 86.7% correct, CI[77%, 100%], p <.01). Children only considered speed as an indicator of competence when an agent reached her goal (Exp.2 (preregistered): n = 20; 90% correct, CI[80%, 100%], p <.01). When agent-based speed was matched, 4-5-year-olds successfully used task-based differences in tower difficulty to infer relative competence (Exp.3 (preregistered): n = 20; 95% correct, CI[90%, 100%], p <.01), which we replicated via online testing (n = 20, 85%, CI = [70%, 100%], p <.01). Given such robust success, Experiments 4-5 tested whether children can integrate information about task-based and agent-based features in the absence of end-state perceptual cues by having children judge the relative competence of two agents who took the same amount of time to make towers that looked identical in the end, but differed in their relative building difficulty (e.g. one tower started with more blocks assembled in Exp 5). Although 4-6-year-olds performed at chance (Exp.4: n = 20, 55% correct, CI [35%, 75%]), expanding the age range to 4-8-year-olds revealed an effect of age (Exp.5: n = 50 b = .67, p <.01) and success by age 8 (90% correct, p = .02).

These results provide preliminary evidence that a basic understanding of relative competence emerges earlier than previously thought, but that an abstract, adult-like concept of competence may continue to develop throughout childhood.

Authors

• Julia Anne Leonard, Yale University
  Presenting Author

• Grace Bennett-Pierre, Temple University
  Non-Presenting Author

• Teresa Garcia, University of Maryland - College Park
  Non-Presenting Author

• Hyowon Gweon, Stanford University
  Non-Presenting Author