Item analysis of a reading comprehension test using the Rasch model for EFL learners

This study aimed to analyze the quality of a reading comprehension test using Rasch measurement. A quantitative method with a descriptive design was employed, involving 200 eleventh-grade students from a senior high school in Sukoharjo, Central Java, Indonesia. The instrument consisted of 29 multiple-choice reading comprehension items and was piloted from June to July 2025. Data were collected through test administration and analyzed using Quest software based on the Rasch model. The results showed that the instrument demonstrated good reliability, with item reliability of 0.89 and person reliability of 0.87. Item fit analysis indicated that 13 items fit the Rasch model, while 16 items were misfitting and required revision. Most items were classified as having a medium level of difficulty (93.10%), and the majority of students were in the high reading comprehension ability category (70%). Overall, the instrument showed satisfactory measurement quality; however, several items need refinement to improve the validity and precision of the test. This study provides empirical support for the utility of Rasch analysis in test evaluation and offers a basis for future improvement of reading comprehension assessments.