<![CDATA[Conventional physics assessments often emphasize final numerical answers and therefore provide limited information about where students' reasoning breaks down during problem solving. This study developed and preliminarily validated a process-oriented assessment instrument for heat learning. A quantitative preliminary instrument-validation design was used. Four contextual essay problems were decomposed into five stages focusing on the problem, describing the physics, planning the solution, executing the plan, and evaluating the answer yielding 20 scored sub-items using a 0–4 analytic rubric. Five experts evaluated content validity using I-CVI, S-CVI/Ave, and S-CVI/UA, followed by an empirical trial with 30 Grade 12 students. Empirical analyses included corrected item-total correlations, Cronbach's alpha, item difficulty, and item discrimination. The instrument obtained S-CVI/Ave = 0.960 and S-CVI/UA = 0.800, with I-CVI values ranging from 0.80 to 1.00. Corrected item-total correlations ranged from 0.605 to 0.909, Cronbach's alpha was 0.975, difficulty indices ranged from 0.583 to 0.750, and discrimination indices ranged from 0.406 to 0.750. These findings provide preliminary evidence that the instrument can support stage-specific diagnosis of students' problem-solving performance in heat learning. Larger-sample validation and inter-rater reliability studies are still required.]]>
Copyrights © 2026
