The sources of children's source errors in judging causal inferences

Understanding the source of an inference is critical for regulating inferences and understanding mental process. Four experiments determine if and why young children make source errors in attributing causal inferences to stories rather than to themselves. First, second, and fourth graders listened to short stories containing an inconsistent goal and outcome that invited a causal inference. An inference was probed after each story and then attributed by the subject to self or the story. In Experiments 1 and 2, clues about the inference and the inference probe at test were manipulated to test the representation hypothesis that source errors vary with the likelihood that a story representation in memory contains causal information that matches retrieval cue information. The results were consistent with the hypothesis for all grades. In Experiments 3 and 4, instructions and test delay were manipulated to show that differential access to verbatim information in a story representation contributes to developmental differences in source errors. The results have implications concerning children's inference monitoring and developing theory of mind and about the developmental relation between memory for verbatim information and inference.