The demands of 21st-century learning emphasize critical thinking skills as part of the 4Cs, requiring assessment instruments capable of revealing students' thinking processes more authentically in electrochemistry material. Based on research conducted at SMAN 3 Semarang in November 2025 involving 86 grade XII students, it was found that assessment was still traditional, using paper-based written tests, and was still subjective. The quantitative method and correlational design used in this study focused on testing assisted storytelling tests by means of video recording and transcription with the help of artificial intelligence (AI) to measure critical thinking skills based on Facione’s indicators (interpretation, analysis, inference, evaluation, and explanation). The instrument’s feasibility was reviewed through content validity by two validators and analyzed using Gregory's formula, while reliability was tested using Cronbach's Alpha. The results showed a content validity coefficient of 0.80 (very high category), indicating that the instrument was valid in terms of content. The instrument's reliability obtained an α = 0.623, which is considered sufficiently reliable. Students’ critical thinking skills tend to be high in Interpretation–Analysis–Inference, while Evaluation and Explanation are relatively lower. Categorically, based on total scores, 88.37% of students are in the very critical category, 10.47% in the critical category, and 1.16% in the fairly critical category. Although the results show that the relationship between Critical Thinking indicators tends to be weak.
Copyrights © 2026