Nurussaniah Nurussaniah, Punaji Setyosari, Dedi Kuswandi, Saida Ulfa
The accurate assessment of analytical thinking in physics, particularly in magnetism, poses substantial challenges due to the limitations of conventional tools in measuring higher-order cognitive skills. This study aimed to validate an analytical skills test in physics, based on Bloom’s Revised Taxonomy, with an emphasis on the dimensions of differentiating, organizing, and attributing within the subject of magnetism. The assessment instrument, comprising 15 multiple-choice items, was administered to a sample of 84 twelfth-grade students. A psychometric evaluation was conducted utilizing the Rasch Model, which included analyses of item fit, reliability, unidimensionality, and gender fairness through Differential Item Functioning (DIF). The results indicated strong internal consistency, with a Cronbach’s alpha of .91, and high item reliability at .94. Furthermore, the item-person distribution showed a well-aligned outcome. Principal Component Analysis (PCA) confirmed the unidimensionality of the test, while the DIF analysis revealed several items exhibiting gender-related response patterns, thereby suggesting a need for refinement to achieve fairness. The findings endorse the application of this validated instrument for assessing higher-order analytical thinking in the field of physics and highlight the significance of equitable assessment practices within STEM education. © 2025, Scientia Socialis Ltd. All rights reserved.
Department of Educational Technology, State University of Malang (Universitas Negeri Malang), Jl. Semarang 5, East Java, Malang, 65145, Indonesia
