Analisis Kualitas Soal Sumatif Akhir Semester (SAS) Mata Pelajaran IPAS Kelas IV

Abstract

With the implementation of the Merdeka Curriculum, which emphasizes competency-based assessment, the quality of evaluation questions becomes a crucial aspect in measuring student learning outcomes. This study aims to analyze the quality of the End-of-Semester Summative Assessment (SAS) questions for the Natural and Social Sciences (IPAS) subject in Grade IV at SDN Wirowongso 2, and their relation to students' competency achievement. This research employs a descriptive method combining both qualitative and quantitative analyses. The instruments analyzed include multiple-choice, short answer, and essay questions based on aspects such as validity, reliability, difficulty level, discriminating power, and fulfillment of HOTS indicators according to Bloom’s taxonomy. The results show that out of 25 multiple-choice questions, 84% were deemed valid and 16% invalid, although overall they were not reliable (coefficient 0.769). For the short answer questions, 70% were valid and 30% invalid, with a reliability coefficient of 0.603. All essay questions were valid (100%) but still did not meet the reliability criteria (0.754). In terms of difficulty level, there was a range from very difficult to very easy questions, with a dominance in the easy category. The discriminating power was mostly low, with only 5 questions categorized as fair, and 3 questions having no discriminating power at all. Furthermore, the cognitive level distribution of the questions was still dominated by C1 (remembering) and C2 (understanding), with very few HOTS-oriented questions. In conclusion, the quality of the Grade IV IPAS SAS questions still needs improvement, especially in terms of discriminating power, reliability, and fulfillment of higher-order thinking aspects to more accurately measure students’ competency achievement