PENGARUH KESALAHAN PENGUKURAN TERHADAP KREDIBILITAS EVALUASI PENDIDIKAN AGAMA ISLAM (PAI): STUDI PADA INSTRUMEN TES DAN NON-TES di SDN 023 TENGGARONG

Abstract

Measurement errors are factors that can reduce the accuracy and quality of learning evaluation results. In the context of Islamic Education (PAI) evaluation at SDN 023 Tenggarong, such errors may occur in cognitive, affective, and psychomotor assessments, both in test and non-test instruments. This study employed a literature review method, examining evaluation theories, psychometric principles, and research on assessment practices in primary schools. The findings indicate that measurement errors mainly stem from instruments that do not meet validity and reliability standards, inaccuracies in item construction, teacher subjectivity in observations, rating bias, and administrative constraints. Internal learner factors—such as emotional state, motivation, and understanding of instructions—also influence scores, causing them to not always reflect actual ability. The study concludes that improving evaluation credibility requires professional accuracy from teachers and a quality-oriented evaluation culture supported by continuous training, reflective practice, and academic supervision. Implementing standardized evaluation procedures is expected to produce more reliable data and support accurate instructional decision-making