Scientific reasoning involves a person's ability to think and act in ways that help advance their understanding of the natural world. Young children are naturally inclined to engage in scientific reasoning and display an emerging competence in the component skills of, for example, hypothesizing, experimenting and evaluating evidence. Developmental psychology research has shown that same-age children often differ considerably in their proficiency to perform these skills. Part of this variation comes from individual differences in cognition; another part is due to the fact that the component skills of scientific reasoning emerge at a different age and mature at a different pace. Significantly less attention has been paid to children's capacity to improve in scientific reasoning through instruction and deliberate practice. Although elementary science lessons are generally effective to raise the skill level of a group of learners, not all children benefit equally from the instructional treatment they receive. Knowing what causes this differential effectiveness is important as it can inform the design of adaptive instruction and support. The present study therefore aimed to identify and explain how fifth-graders (N = 138) improve their scientific reasoning skills over the course of a 5-week inquiry-based physics unit. In line with our expectations, significant progress was observed in children's achievements on a written scientific reasoning test, which was administered prior to and after the lessons, as well as in their responses to the questions and assignments that appeared on the worksheets they filled out during each lesson. Children's reading comprehension and mathematical skillfulness explained a portion of the variance in children's pretest-posttest gain. As these overall results did not apply equally to all component skills of scientific reasoning, we recommend science teachers to adapt their lessons based on children's past performance in reading and math and their actual performance of each scientific reasoning skill. The orchestration and relative effectiveness of both adaptive science teaching approaches is an interesting topic for future research.
