Reading focus
Syllabus: How Do Children Develop Language and Thought?
Use the controlled demonstrations to connect the structure of evidence with exploratory action. Apply the earlier distinction between an observed response and an inferred mental process to the claim that play can be informative.
A question to consider. Which action would separate two possible causes, and what further measure would show whether the child learned from performing it?
Why this reading
Exploration may depend on what is still uncertain, not just what is new. Schulz and Bonawitz vary the evidence children receive about a toy’s causal workings and then observe their play. The study connects the information available to a learner with the actions that could make it more informative.
Before you read
Evidence is confounded when potential causes vary together so that their separate contributions cannot be determined. An intervention deliberately changes one factor. A preference for a toy, time spent exploring it, and an informative action are different measures. Novelty refers to unfamiliarity, which need not be the same as unresolved causal information.
Read for
- Reconstruct the demonstration conditions. Explain why the unconfounded comparisons address alternatives involving the number of demonstrations, observed effects, and opportunities to act.
- Compare play with the familiar and novel toys. Separate the measures of preference from the analysis of actions that could isolate a cause.
- Identify the strongest causal claim the manipulation supports. What evidence would be needed to show that children learned the correct causal relation during their exploration?
Consider afterwards
- Two controls always change together when a toy activates. Propose an informative action, then explain how you would distinguish a child’s deliberate attempt to resolve uncertainty from an action that happens to provide information.
Editorial context
The experiment supports a relation between unresolved causal evidence and exploration in a particular preschool task. It does not directly assess children’s causal knowledge after play or demonstrate an explicit scientific method in every child. Its small condition groups and restricted sample leave questions about reliability and broader settings. The controlled comparisons are useful precisely because they distinguish uncertainty from several simpler explanations of play; a general recommendation about discovery learning would require additional evidence.