Scientific reasoning in young children: Preschoolers' ability to evaluate covariation evidence

Preschool children's basic scientific reasoning abilities were investigated in two experiments. Consistent With findings by Ruffman et al. (1993), Experiment I showed that even 4-year-olds can evaluate patterns of covariation evidence. However, even 6-year-olds had difficulties interpreting non-covariation evidence. Experiment 2 showed that 5-year-olds could overcome this difficulty when prompted to expect no causal relationship between two variables. Experiment 2 further showed that preschoolers' evidence evaluation skills were affected by their pre-existing causal beliefs. However, their performance was above chance even when the evidence contradicted a prior belief they held with some conviction. In sum, our results demonstrate a basic understanding of the bypothesis-evidence relationship in preschool children, thus contributing to a revision of the picture of the scientifically illiterate preschooler.