<![CDATA[Assessment quality is a crucial element in ensuring that science education achieves its intended learning objectives, particularly at the junior secondary level where foundational concepts are introduced. This study aimed to evaluate the quality of teacher-constructed mid-term examination items in Grade VII science by focusing on item difficulty, discrimination index, and overall reliability. Using a quantitative descriptive design, data were collected from 68 student responses to 25 test items, comprising 20 multiple-choice and 5 essay questions, and analyzed through classical test theory. The findings showed that 28% of items were classified as easy, 52% as moderate, and 20% as difficult, with 60% demonstrating acceptable or good discrimination power and a KR-20 reliability coefficient of 0.72, indicating adequate internal consistency. While the results suggest that the test achieved a balanced level of difficulty and acceptable reliability, the presence of items with poor discrimination and extreme difficulty levels reveals weaknesses in test construction. The discussion underscores that systematic item analysis is essential to refine teacher-made assessments and align them with both curriculum standards and international benchmarks. The novelty of this study lies in its focus on teacher-constructed science tests in junior secondary schools in Indonesia, a context that remains underexplored in the literature. The implications of this research point to the need for enhanced teacher assessment literacy, institutional support, and continuous evaluation practices to improve the validity and reliability of classroom-based examinations.]]>
Copyrights © 2023
