Adapting scientific reasoning instruction to children's needs: effects on learning processes and learning outcomes

Scientific reasoning helps children understand the world around them. Teaching scientific reasoning can be challenging because not all component scientific reasoning skills develop at the same age and not all children learn these skills at the same pace. Adaptive support thus seems called for. We designed two types of adaptive instruction, based on children's standardised test scores (macro-adaptive; n = 58) or their performance in the previous lesson (micro-adaptive; n = 46), and tested their effectiveness against a non-adaptive control condition (n = 49). Analysis of pre- and post-test scores showed comparable improvements in all three instructional conditions. As many children in both adaptive conditions received medium support, additional analyses were done on children in the macro-adaptive condition who received high or low-support worksheets, and their control group counterparts. Learning gains for these groups were similar. Children's overall task performance during the lessons also improved, and this improvement interacted with condition. These results suggest that more specific information on children's performance and more frequent and precise adaptations might lead to better learning outcomes. As this was not possible in this study, future research should explore hybrid solutions that enable children to practice scientific reasoning with physical materials while receiving adaptive support via their computers.