Penyusunan Instrumen Evaluasi Pembelajaran: Analisis Pembuatan Soal Pilihan Ganda dan Esai/Uraian

Abstract

An empirical problem frequently found in educational practice is the low quality of learning evaluation instruments, which leads to inaccurate measurement of students’ learning outcomes. This study aims to analyze the development of learning evaluation instruments through an examination of multiple-choice and essay questions as the most commonly used test formats in schools. The research employed a qualitative library research method by systematically reviewing textbooks, journal articles, and scholarly documents related to educational evaluation. The findings indicate that high-quality evaluation instruments must meet the principles of validity, reliability, objectivity, and practicality. Multiple-choice items are effective for objectively assessing a broad range of content and allow empirical analysis through difficulty and discrimination indices, while essay items are more effective in assessing higher-order thinking skills but require detailed scoring rubrics to minimize subjectivity. The study concludes that teachers’ mastery of the principles and procedures of test construction plays a crucial role in ensuring accurate assessment results. The implication of this study highlights the need to strengthen teachers’ competencies in developing evaluation instruments to improve the quality, fairness, and effectiveness of learning assessment.