<![CDATA[Improving the quality of basic education requires evaluation instruments capable of objectively and accurately measuring learning outcomes. However, in practice, the quality of multiple-choice questions in elementary schools still does not fully meet the criteria for good evaluation, requiring an in-depth analysis of the questions used. This study aims to analyze the quality of multiple-choice questions on natural resources for fourth-grade students at SD Negeri Inpres Kota Baru Nabire by examining validity, reliability, discriminatory power, and difficulty levels. The method used was a quantitative approach with a descriptive evaluative approach, involving fourth-grade students as research subjects. Data collection techniques included tests and data analysis using the Anates application. The results showed that all questions were valid with correlation coefficients above r table. The instrument's reliability was in the good category with a Cronbach's Alpha value of 0.731. Most questions had sufficient to good discriminatory power, and the difficulty level was in the easy and medium categories, indicating a balanced distribution. These findings provide an important contribution to the development of higher-quality learning evaluation instruments based on empirical data. The study's conclusions indicate that the test instrument is suitable for use with some improvements to items with low discriminating power. It also recommends developing items based on Higher Order Thinking Skills and further analyzing the effectiveness of distractors to improve the quality of future learning evaluations.]]>
Copyrights © 2026
