No abstract

Objectives The purpose of this study is to conduct action research on a mathematics class for students with below basic academic ability using an AI-based adaptive learning system to which a personalized instruction strategy is applied and to explore educational changes between learners and teachers in this process. Methods For this reason, teachers and researchers conducted math classes using an AI-based adaptive learning system that applied personalized instruction strategies derived for middle school students with the achievement of below basic proficiency level for about four months. Collecting main data was from observation logs, reflection logs, interview, learning data, and mathematical learning affective domain test, and the data were analyzed as the mixed research method. Results During research, a class for students with below basic academic ability was changed to a personalized class using the AI-based adaptive learning system, and in this process, planning, execution, and reflection of personalized class using AI-based adaptive learning were analyzed. The learners who participated in this study were diagnosed and recovered the prerequisite learning deficits and the cognitive domains generally increased, while no significant differences were found in the affective domains. Through qualitative analysis, we confirmed that mathematical interest and mathematical self-efficacy improved and learning satisfaction increased. Teachers realized personalized classes, experienced roles of new instructors, and discovered the possibility of personalized and customized instruction within the school. Conclusions Based on the results of this study, the implementation process, advantages, and limitations of classes using the AI-based adaptive learning system applied with personalized instruction strategies were reviewed, and measures were discussed to realize personalized instruction at schools.