<![CDATA[This study addresses the lack of specialized diagnostic instruments capable of pinpointing specific cognitive failures in physics data literacy, particularly within Thermodynamics. To bridge this gap, a Distractor-Based Cognitive Diagnostic Test (DB-CDT) was developed and validated for eleventh-grade high school students in Yogyakarta. Utilizing a Research and Development (R&D) approach with the ADDIE framework, the instrument was built upon the DINA (Deterministic Input, Noisy-AND Gate) cognitive diagnostic model. Expert validation by three specialists yielded an excellent average V-Aiken value of 1.00. Empirical testing on a sample of 540 students across high, medium, and low-tier schools confirmed high psychometric quality: an average INFIT MNSQ of 1.00 (within the 0.77–1.30 range), item difficulty between -2 and +2, and high reliability (0.92). DB-CDT distinguishes itself by using a Q-Matrix to link specific distractors to cognitive errors, such as failures in sign conversion or process interpretation. Analysis of student profiles revealed that the experimental group—utilizing targeted diagnostic feedback—achieved superior mastery in analyzing (96.6%) and interpreting data (90%), whereas the control group, receiving conventional assessment, struggled significantly with inference and synthesis. Theoretically, this research advances the application of Cognitive Diagnostic Models in physics education. Practically, it provides educators with a sensitive tool to profile individual student weaknesses, allowing for more focused instructional interventions in complex scientific topics.]]>
