Nusantara (Jul 2024)
Analisis Akurasi Soal Pilihan Ganda Materi Rangka Manusia pada Ulangan Harian Siswa di Sekolah Dasar
Abstract
Purpose – This study aims to determine the quality of multiple choice items on human skeletal material in terms of validity, reliability, difficulty level, and differentiating power. Method – This research uses a quantitative descriptive approach. the subjects in this study were grade V students of State Elementary School 1 Sungai Bungin, totaling 22 students. the sample selection technique used saturated sampling technique. the data collection technique was carried out by the documentation method from the test results. the collected data were analyzed using the SPSS IBM Statistics 27 application to determine the criteria for the items. Findings – This study provides a detailed description of the validity, reliability, differentiating power, and difficulty of the questions. It found 17 valid items (68%), with high reliability (Cronbach's Alpha: 0.798 and 0.792). A total of 4 items had sufficient differentiating power, 19 items were good, and 2 items were excellent. The level of difficulty of the questions varied: 4 difficult questions (16%), 13 medium questions (52%), and 8 easy questions (32%). Research Implications – This research has important implications for improving the quality of education in elementary schools. The results of analyzing the accuracy of multiple choice questions on human skeletal material can be used to improve and develop questions that are more appropriate in measuring student knowledge. Teachers can adjust teaching methods based on the difficulty level of the questions, improve learning effectiveness, and provide more constructive feedback. In addition, the findings can assist in more targeted curriculum development and teacher training in the creation of high-quality questions. Thus, this study contributes to the improvement of student learning outcomes and the implementation of data-driven evaluation in the primary education system.
Keywords
