Applying the Rasch Model to Assess Retention and Transfer Test Instruments in Science Education on Additive and Addictive Substances

Abstract

Purpose of the study: This study aims to evaluate the quality of items in retention and transfer tests related to additive and addictive substances. Using Rasch modeling, the study seeks to enhance the management of learning evaluations and improve our understanding of student abilities and question quality. Methodology: The research utilizes the Rasch Model to analyze retention and transfer test instruments on science topics involving additives and addictive substances. Conducted with Winstep software, the analysis focuses on the performance of 92 purposively sampled 8th-grade students during their first semester of junior high school. The study examines retention and transfer abilities, comprehensively evaluating the test items. Main Findings: The Winstep program analysis reveals that, according to the Rasch model, the average ± MNSQ Outfit values for both items and persons are 0.92. The Outfit ZSTD values for items and persons are -0.12 and -0.01, respectively. The instrument's reliability, measured by Cronbach's alpha, is 0.60, indicating moderate reliability. The research findings demonstrate that each item in the instrument is valid and reasonably reliable, with all 20 items deemed suitable for assessing student performance in retention and transfer tests. Novelty/Originality of this study: This study offers a detailed examination of retention and transfer test instruments' quality using the Rasch Model, providing valuable insights for enhancing the accuracy and reliability of these assessment tools. The research significantly improves educational assessments in science education, particularly in evaluating students' understanding of additives and addictive substances.