Development of an Instructional Design Model for Co-designing Science Inquiry Teaching with Generative AI Shin Ju Ran Advisor : Prof. Park Hyun Ju, Ph.D. Department of Science Education, Graduate School of Chosun University The purpose of this study w...
Development of an Instructional Design Model for Co-designing Science Inquiry Teaching with Generative AI Shin Ju Ran Advisor : Prof. Park Hyun Ju, Ph.D. Department of Science Education, Graduate School of Chosun University The purpose of this study was to develop an instructional design model for co-designing science inquiry teaching with generative artificial intelligence (AI). Generative AI was conceptualized as a collaborative partner that supports teachers’ instructional design while fostering stuedents’ inquiry thinking. Based on this perspective, the Generative AI-Integrated Science Inquiry Design (GAI-SID) model was developed. The study employed a two-stage research design consisting of preliminary analysis and model development. The preliminary analysis included an analysis of 172 student science inquiry project reports from 2018 and 2025, a survey of 10 science teachers to validate the interpretations of the observed changes in inquiry errors, a survey of 19 science teachers to investigate the potential uses and concerns of generative AI at each inquiry stage and the appropriateness of the proposed conceptual model, and a survey of 62 pre-service teachers to examine the perceived importance of and readiness for digital-based teaching competencies. In the second stage, an initial GAI-SID model was developed from the findings of the preliminary analysis and iteratively refined through a focus group interview with four science teachers and pilot testing with 12 pre-service teachers. The results were as follows. First, errors related to basic inquiry design, such as formulating research questions, identifying variables, and selecting equipment and materials, tended to decrease in student science inquiry project reports. In contrast, difficulties in higher-order inquiry processes, including data interpretation, scientific reasoning, conclusion drawing, and the appropriate use of references, continued to occur. Science teachers recognized the potential of generative AI to support literature review, inquiry question development, experimental design, feedback, and report writing, while expressing concerns about students’ uncritical acceptance of AI-generated responses and excessive reliance on AI during data interpretation and conclusion drawing. Pre-service teachers regarded digital-based teaching competencies as highly important but reported relatively low readiness to implement them. Second, the preliminary analysis informed a set of instructional design principles that emphasize stage-specific integration of generative AI, iterative refinement of instruction, appropriate allocation of teacher and AI roles, prevention of AI overdependence, and process-oriented assessment. These principles highlight that the scope and role of generative AI should vary across inquiry stages, while students’ scientific judgment in data interpretation, reasoning, and conclusion drawing should remain central. Accordingly, assessment should address not only inquiry products but also inquiry processes, students’ critical evaluation of generative AI outputs, and the evidence supporting their judgments. Third, the final GAI-SID model comprises three stages-Preparation, Design, and Implementation— and ten instructional procedures connected by iterative feedback loops that enable continuous refinement of instructional decisions before, during, and after inquiry teaching. In conclusion, the GAI-SID model provides a systematic framework for co-designing science inquiry teaching with generative AI. It supports teachers in determining the purpose, scope, and timing of AI use across inquiry stages while preserving students’ responsibility for scientific reasoning and enabling continunous improvement of instructional design through iterative feedback. Key words: Generative AI, science inquiry teaching, instructional design model, student science inquiry project reports, GAI-SID