Assessment of Scientific Reasoning: the Effects of Task Context, Data, and Design on Student Reasoning in Control of Variables.

Scientific reasoning is an important component under the cognitive strand of the 21st century skills and is highly emphasized in the new science education standards. This study focuses on the assessment of student reasoning in control of variables (COV), which is a core sub-skill of scientific reasoning. The main research question is to investigate the extent to which the existence of experimental data in questions impacts student reasoning and performance. This study also explores the effects of task contexts on student reasoning as well as students' abilities to distinguish between testability and causal influences of variables in COV experiments. Data were collected with students from both USA and China. Students received randomly one of two test versions, one with experimental data and one without. The results show that students from both populations (1) perform better when experimental data are not provided, (2) perform better in physics contexts than in real-life contexts, and (3) students have a tendency to equate non-influential variables to non-testable variables. In addition, based on the analysis of both quantitative and qualitative data, a possible progression of developmental levels of student reasoning in control of variables is proposed, which can be used to inform future development of assessment and instruction.