Improving Measurement Accuracy and Efficiency of Computerized Adaptive Testing Using Prior Information

Objectives This study investigates the accuracy and efficiency of Computerized Adaptive Testing (CAT) by utilizing prior information to predict students' basic academic levels. Methods To achieve this, a simulation study was conducted with varied conditions, including initial item, partic-ipant groups, and test stopping rule. Results Results indicate that selecting initial items based on predictors of basic academic levels significantly im-proved the accuracy and efficiency of CAT measurements, with the most substantial improvements observed in low-achieving groups. Conclusions The findings suggest the need for further exploration of the applicability of basic academic level pre-dictors across different educational contexts to identify strategies for improving the precision and efficiency of CAT systems.