<![CDATA[This study aims to validate a research‐skills assessment instrument for pre-service science teachers using Rasch modeling to evaluate reliability, item fit, and unidimensionality. The instrument was designed to measure the extent of research skills among undergraduate students in science education programs. A cross-sectional survey was conducted with 114 participants recruited through quota sampling from several universities in Riau, Indonesia. Respondents completed a 40-item research-skills instrument. Data were analyzed with the Rasch model using Winsteps version 3.73. The results indicate satisfactory internal consistency (Cronbach’s alpha = 0.75), excellent item reliability (0.97), and acceptable person reliability (0.75). Rasch diagnostics further support the unidimensionality of the instrument, indicating that the items predominantly measure a single latent construct of research skills. These findings provide evidence for both the reliability and construct validity of the instrument within this population. Practically, the validated instrument can be used by higher education institutions to diagnose and monitor the research competencies of pre-service science teachers and to evaluate the effectiveness of research-oriented instructional interventions. By employing Rasch analysis, this study enhances the precision and defensibility of measurement in the context of research-skills assessment in Indonesia, offering a foundation for ongoing refinement and wider application in similar educational settings]]>
Copyrights © 2025
