Innovation Learning on The Topic of Sound: Rasch Analysis on Team Game Tournament with Uno Card Media

This study aims to analyze the characteristics of the Teams Game Tournament (TGT) learning model instrument using the Rasch model on the topic of sound, incorporating Bloom's cognitive features. The analysis focuses on validity, reliability, discrimination capacity, and difficulty level to determine the instrument's feasibility. The study adopts the classroom action research paradigm. The instrument comprises ten multiple-choice questions. The findings indicate that the instrument demonstrates adequate validity within an acceptable range. The reliability analysis yields a Cronbach's alpha score of 0.82, indicating high reliability, with item reliability at 0.36 and person reliability at 0.32. Additionally, the discriminating power meets the criteria, as shown by Infit Mean Square (MNSQ), Outfit Z-Standard (ZSTD), and Point Measure Correlation (PT Measure Corr). Regarding difficulty level, the instrument includes five questions categorized as difficult, three as moderate, and two as easy.