LAUSR

This study examines progression trends of Chinese students’ scientific reasoning skills across grade levels from elementary school to university. A large-scale survey using the Classroom Test of Scientific Reasoning (CTSR) was conducted with 2669 Chinese students at 13 grade levels (grades 4–16). The construct validity of the CTSR was first examined using Rasch analysis to verify that the various reasoning sub-skills targeted by the test (hypothetical-deductive reasoning, proportional reasoning, correlation reasoning, probabilistic reasoning, and control of variables) form a cohesive cognitive construct. Based on an affirmative result, we proceeded to fit grade-level averages to the best-fit logistic regression model and generated trend lines for the reasoning sub-skills measured by the CTSR. Results show that the cross-grade progression trends generally follow continuous paths through increasingly greater levels of improvement. While all sub-skills display a noticeable growth during middle and high school years, they progress across the grades with different rates. Specifically, the development of proportional reasoning is fast-paced and large, but the progression of hypothetical-deductive reasoning and control of variables is markedly tardy and small. Formal science instruction, particularly the emergence of multiple domain-specific science courses in middle and high school, can influence the cross-grade progression trends. Possible causes and future improvement in science education are proposed. This study contributes to our understanding about the development of scientific reasoning and offers important implications for science instruction and research at different grades.