<![CDATA[This study aims to evaluate the validity and reliability of a science literacy instrument based on the Rasch Model in measuring students' abilities on socio-scientific issues. The analysis was conducted on 20 items using the Winsteps software. The results indicate that the instrument demonstrated good reliability, with an item reliability value of 0.96 and a person reliability value of 0.65. The Wright Map revealed that most items were well-aligned with the respondents’ abilities, although some items fell outside the majority's ability range. The analysis of item difficulty showed adequate variation, with the most difficult item at logit 2.09 and the easiest at logit -2.49. However, the DIF analysis identified bias in item 20M, which exhibited significant differences between male and female groups. These findings highlight the importance of revising specific items to enhance fairness in measurement. This study provides significant contributions to the development of educational assessment instruments based on the Rasch Model, ensuring validity, reliability, and fairness in evaluation. The instrument is expected to serve as an effective tool to support science literacy-based learning across various educational levels.]]>
Copyrights © 2024
