지능형 과학실 ON 환경 기반 AI 과학 탐구수업모형 개발 및 탐색적 적용

초록

This study develops AI-based science inquiry instructional models to address stepwise cognitive challenges and limitations of individualized teacher support in science inquiry education. Within a Design-Based Research (DBR) framework, five textbook-based inquiry models and a four-stage self-directed inquiry model are developed based on five design principles derived from literature analysis, focus group interviews with leading teachers, and textbook inquiry analysis. The models are applied twice to sixth-grade elementary and first- and third-grade middle school students. In the second application, the AI-based group shows statistically significantly higher performance in the observation element for sixth graders (t = 8.45, p < .001), directionality of variable relationships for first-grade middle schoolers (t = 2.32, p = .025), and counterargument recognition/rebuttal for third-grade middle schoolers (t = 2.19, p = .033). This study is significant in presenting inquiry instructional models that systematically integrate AI support based on inquiry type and stepwise cognitive functions, with an exploratory empirical examination of their educational effectiveness limited to the observational, problem-finding, and argumentation inquiry contexts in which the models were applied.

키워드