earticle

영어

The purpose of this study is to analyze College Scholastic Ability Test items based on the Rasch Model, identify items that do not fit, and describe difficulty levels compared to student levels. Winsteps software, which allows users to analyze dichotomous data for investigating problems that are inherent in test items, is used for statistical analysis. The findings of this study reveal that 7 out of 28 items do not follow the assumptions of the Rasch Model. The analysis of item difficulty and item map shows that most of the 387 subjects are unable to answer some item types appropriately due to ambiguous distractors that affect item validity. Given the presence of misfit items and items of extreme levels of difficulty, suggestions for eliminating ambiguity, improving readability, and maintaining the degree of difficulty in designing reading test items are made.