Background: Misconceptions in geometric optics remain persistent, including among prospective physics teachers, and are often difficult to detect using conventional assessments that focus only on final answers. Although four-tier diagnostic tests provide deeper conceptual information, their implementation is generally static and may reduce assessment efficiency. Aims: This study aimed to develop and evaluate a rule-based adaptive four-tier diagnostic quiz system capable of identifying misconceptions more efficiently while maintaining stable diagnostic classifications. Method: The study employed a research and development approach using the ADDIE model. A web-based adaptive diagnostic was developed by integrating a four-tier scoring scheme with transparent rule-based adaptive decisions. The platform was tested on 36 prospective physics teacher students. Data were analyzed descriptively to examine adaptive test length, diagnostic pathways, and classification stability. Results: The adaptive system reduced the average test length from 15 static items to 11.5 items, representing an efficiency gain of 23.3%. A total of 77.8% of participants reached diagnostic stability before the maximum item limit, and the classification consistency rate reached 83.3%. The system also revealed variations in misconception patterns across topics, with concave mirror concepts showing the highest proportion of strong misconceptions. Conclusion: The rule-based adaptive four-tier system improved diagnostic efficiency while maintaining stable classification outcomes. The transparent adaptive mechanism makes the system suitable for formative diagnostic assessment in physics education, although further studies with larger samples are recommended.
Copyrights © 2026