Analysis of Arabic Language Madrasah Assessment Test Items Based on Educational Assessment Standards at Al-Ikhlas Nanggulan Islamic Senior High School

Abstract

Analyzing test items is essential for improving the quality of the Arabic language curriculum, as it helps identify weaknesses in the evaluation process. This study aims to examine the validity, reliability, difficulty level, and discrimination power of Arabic language questions used in the Madrasah Assessment at MA Al-Ikhlas Nanggulan for the 2022–2023 academic year. A descriptive quantitative method was employed, with 40 multiple-choice questions as the research population. Data were collected through documentation of the answer sheets of 25 twelfth-grade students. Data analysis included the Product-Moment correlation for validity, Cronbach’s Alpha for reliability, the difficulty index (P), and the discrimination index (D). The results showed that the test items were generally of good quality. Of the 40 items, 65% were valid, while 35% were invalid. The reliability coefficient reached 0.915, indicating very high reliability. In terms of difficulty, 17.5% of the items were easy, 70% were moderate, and 12.5% were difficult. Regarding discrimination power, 62.5% of the items were classified as good, 17.5% as moderate, 12.5% as poor, 5% as ineffective, and 2.5% as negative. These findings indicate that item analysis is an effective tool for evaluating and improving the quality of Arabic language assessments.