Item analysis of high school English test items using Item Response Theory

This study employed Item Response Theory to improve specific aspects of examination questions, such as item difficulty, discrimination, and reliability, formulated by high school teachers. Utilizing the jamovi statistical software, the Rasch model was applied to analyze the multiple-choice questions from the final exams of 247 first-year high school students in a G metropolitan city. The item analysis revealed that the exam exhibited satisfactory person reliability. Except for two items, the difficulty levels and scoring were appropriately aligned with the students’ capabilities. Additionally, apart from one item, the difficulty and discrimination of the other items were found to be adequate. Notably, fill in the blank type item was identified as favoring female students through Differential Item Functioning analysis, prompting a in-depth discussion on potential modifications. This research provides objective indices through the Rasch model to evaluate the reliability, difficulty, discrimination, and differential item functioning of multiple-choice questions, thereby expected to enhance teachers’ assessment skills.