Effects of Parameter-Based AI Composition on Self-Efficacy and Interest in Elementary Music Education

This study empirically examined the effects of a 'modular composition class' using the parameter-based generative AI tool MusiaOne on elementary students' musical self-efficacy and interest. The program was implemented with 70 fifth-grade students at an AI-leading elementary school in the Seoul metropolitan area and consisted of eight sessions designed upon the GATe (Generative AI Teaching and learning) framework. A mixed-methods approach was employed, combining quantitative analyses of pre- and post-tests with qualitative data from observations, worksheets, teacher interviews, and student reflections. The results revealed statistically significant improvements in both self-efficacy and interest, and students increasingly engaging in creative music-making. These findings suggest that AI-assisted modular composition can be an effective pedagogical strategy for fostering autonomy and creativity in music education. In particular, it highlights the potential for students to experience music creation without a heavy theoretical burden, supporting the validity and scalability of popular music education in public school curricula.