Assessing Educational Instrument Validity-Reliability and Student Performance: Rasch Model Insights

This study aims to assess the validity-reliability and student performance. A cross-sectional research design with a quantitative method was employed in this study. The participants of this research were 34 students divided into 12 males (35.3%) and 22 females (64.7%) with an age range of 16-17 years. The data collection used a Quizizz with 15 questions online but still under teacher supervision. The results demonstrate moderate reliability and validity per Rasch analysis, with a person reliability score of 0.63 and an item reliability score of 0.46, indicating a need for item refinement. The person separation index (1.31) and item separation index (0.93) show moderate differentiation, while a Cronbach alpha of 0.70 and raw variance explained by measures (41.6%) support instrument validity. The characteristics of the items and persons reveal that most items align well with the model expectations, although some exhibit variability or inconsistencies that need addressing.