This research developed socio-cultural literacy test items integrated with a deep learning approach for fifth-grade elementary school students. The study involved analyzing potential and problems, collecting data and information, developing a design, revising the design, finalizing the draft, testing, analyzing, and reporting. Validation was conducted by six experts, two each in the material, language, and construction aspects. Validity based on Aiken's V-value for the material aspect was 0.625-1, indicating a moderate interpretation and very high validity. The language aspect had a value of 0.65-0.75, indicating a moderate interpretation, and the construction aspect had a value of 0.75-1, indicating a very valid interpretation. Interrater reliability for the material aspect averaged 79%, classified as strong; the language aspect averaged 89%, and the construction aspect averaged 89%, classified as near perfect. The instrument was administered to 25 fifth-grade students and empirically tested in SPSS version 29. Items 1 and 2 were invalid, items 3-15 were highly valid and very highly valid. The reliability was 0.909, a very high value. The difficulty index for items 1-6 was easy, and items 7-15 were moderate. The discrimination power of items 1 and 2 was poor; item 3 was moderate. Items 4, 8, 9, 10, 12, and 13 were classified as very good, and items 5, 6, 7, 14, and 15 were classified as good. Therefore, based on the empirical test results, the developed instrument met the requirements for content validity and was suitable for use.
Copyrights © 2025