Validation of an AI-based automatic assessment system for scientific literacy instruments on green chemistry integrated with ethnochemistry

This study aims to measure the content validity of an AI-based automatic assessment system for scientific literacy instruments on green chemistry integrated with ethnochemistry using Aiken’s V index. The research design adopts the R&D model with the ADDIE approach, limited to the analysis, design, and development stages. A combination of quantitative and qualitative descriptive methods was utilized through purposive sampling, engaging 10 experts—chemistry education lecturers from UNS, UNY, UM, and high school chemistry teachers in Surakarta. Data were obtained through Focus Group Discussions (FGD), focusing on three aspects: context, language, and construct. Essay-based instrument items were rated using four relevance categories: irrelevant, less relevant, quite relevant, and relevant. Items with an Aiken index ≥ 0.73 were considered valid. The results showed that 1 question item had moderate validity and 14 question items had high validity, making them suitable for further testing. The questions were designed to assess students’ understanding of green chemistry principles and their ability to relate these to ethnochemistry practices—local cultural knowledge that supports sustainability. The formulation of the questions also demonstrated how the application of local culture can be interpreted through green chemistry to enhance scientific awareness and environmental responsibility. Based on the content validation results and subsequent revisions, the instrument is deemed ready to move forward to the limited trial stage.