Contributions of causal reasoning to early scientific literacy.

Although early causal reasoning has been studied extensively, inconsistency in the tasks used to assess it has clouded our understanding of its structure, development, and relevance to broader developmental outcomes. The current research attempted to bring clarity to these questions by exploring patterns of performance across several commonly used measures of causal reasoning, and their relation to scientific literacy, in a sample of 3- to 5-year-old children from diverse backgrounds (N = 153). A longitudinal confirmatory factor analysis revealed that some measures of causal reasoning (counterfactual reasoning, causal learning, and causal inference), but not all of them (tracking cause-effect associations and resolving confounded evidence), assess a unidimensional factor and that this resulting factor was relatively stable across time. A cross-lagged panel model analysis revealed associations between causal reasoning and scientific literacy across each age tested. Causal reasoning and scientific literacy related to each other concurrently, and each predicted the other in subsequent years. These relations could not be accounted for by children's broader cognitive skills. Implications for early STEM (science, technology, engineering, and math) engagement and success are discussed.

Young children's metacognitive awareness of confounded evidence.

Young children selectively explore confounded evidence-when causality is ambiguous due to multiple candidate causes. This suggests that they have an implicit understanding that confounded evidence is uninformative. This study examined explicit understanding, or metacognitive awareness, of the informativeness of different qualities of evidence during early childhood. In two within-participants conditions, children (N = 60 5- and 6-year-olds) were presented with confounded and unconfounded evidence and were asked to evaluate and explain their knowledge of a causal relation. Children more frequently requested further information in the confounded condition than in the unconfounded condition. Nearly half of them referred to multiple candidate causes when explaining confounded evidence. Our data demonstrate that young children can reason explicitly about the informativeness of different kinds of evidence.