Analysis of the Quality of Sudents’ Critical Reasoning Instruments Using Rasch Model

This research aims to analyzes the validity and reliability of a students' critical reasoning questionnaire using the Rasch Model. This quantitative descriptive research employed a non-experimental survey design involving 122 fourth-semester undergraduate students from the College of Teacher Training and Education (STKIP) Melawi. Data were collected through an online questionnaire distributed via Google Forms. Validity was assessed through unidimensionality and item fit analysis, while reliability was measured using person reliability, item reliability, and Cronbach’s Alpha. The results indicate that the instrument has not yet achieved good item validity, with only 4 out of 20 items classified as valid. Reliability analysis showed Cronbach’s Alpha of 0.55 (bad category), item reliability of 0.98 (special category), and person reliability of 0.44 (weak category). Item difficulty analysis identified four difficulty levels: 2 items were very easy, 9 easy, 6 difficult, and 3 very difficult. The Rasch model’s item discrimination analysis produced eight item groups based on an item separation index of H = 9.97. These findings emphasize the need to refine questionnaire items and enhance respondent engagement. A validated instrument will support educational assessments by providing a reliable tool for measuring students’ critical reasoning skills in teacher training institutions.