Momentum: Physics Education Journal
Vol. 10 No. 1 (2026)

RASCH model analysis: validity and reliability of concept understanding test instrument on temperature and heat

Kafa Pramitha Anggraini Indhira Artanti (Universitas Negeri Malang, Indonesia)
Sherly Eka Putri Damayanti (Universitas Negeri Malang, Indonesia)
Shinta Nuriyah Mahbubiyah Royani (Universitas Negeri Malang, Indonesia)
Bintan Nuril Irvaniyah (Universitas Negeri Malang, Indonesia)
Natasya Mustika Elvina Rossa (Junior High School Antartika, Indonesia)
Anggita Meilina Putri (Universitas Negeri Malang, Indonesia)
Khoirun Nisa' (National Dong Hwa University, Taiwan)



Article Info

Publish Date
27 Jul 2026

Abstract

This study aims to analyze the quality of a concept understanding test instrument on temperature and heat using the Rasch Model approach, focusing on validity, reliability, and item difficulty level. This research employed a descriptive method with a quantitative approach. The sample was selected through simple random sampling and involved 75 students of class XI at Senior High School. The research instrument consisted of 15 multiple-choice items covering concepts related to temperature and heat. Data were analyzed using the Rasch Model assisted by Ministep software to examine item fit, person reliability, item reliability, and the distribution of item difficulty. The results showed that all 15 items were valid based on the INFIT and OUTFIT criteria of MNSQ and ZSTD values. The person reliability was categorized as poor, whereas the item reliability was categorized as very good. The item difficulty level was distributed into four categories, consisting of 4 easy items, 2 moderate items, 6 difficult items, and 3 very difficult items. These findings indicate that the Rasch Model is effective in identifying the validity, reliability, and difficulty distribution of test items. The study implies that the developed instrument can be used by physics teachers and researchers as an empirical basis for evaluating students’ conceptual understanding of temperature and heat. However, the low person reliability suggests the need for further refinement by involving a more heterogeneous sample and developing additional items with a broader range of difficulty levels to improve the instrument’s ability to distinguish students’ conceptual understanding.

Copyrights © 2026