Integrating AI-driven adaptive learning and ethnomathematical game design to address persistent conceptual difficulties in LCM and GCD: Development and evaluation of the DAKOGI platform

Abstrak

This study developed and evaluated DAKOGI, an innovative digital game-based learning platform that integrates adaptive learning algorithms with culturally responsive pedagogy grounded in Dakon, a traditional Indonesian game, to address students’ persistent difficulties in understanding the concepts of Least Common Multiple (LCM) and Greatest Common Divisor (GCD). The research employed a development methodology based on the ADDIE model, encompassing the phases of analysis, design, development, implementation, and evaluation. Participants included 30 secondary school students from SMAN 1 Kuta, Indonesia. The evaluation process incorporated expert validation of content, media design, and pedagogical quality, as well as effectiveness testing through a pre-experimental one-group pre-test–post-test design supported by platform usage analytics. The findings indicate that DAKOGI achieved high levels of quality across all evaluated dimensions, demonstrating strong content validity, pedagogical appropriateness, and technological usability. Furthermore, statistically significant improvements were observed in students’ conceptual understanding of LCM and GCD, with mean achievement scores increasing from 42.17% in the pre-test to 81.50% in the post-test. These gains were further supported by substantial learning improvement indicators and a very large effect size, suggesting a meaningful educational impact. Students also reported high levels of practicality, perceiving the platform as accessible, engaging, and supportive of their learning processes. Finally, DAKOGI provides preliminary empirical evidence for a promising model of technology-enhanced mathematics education that combines game-based learning, adaptive personalization, and ethnomathematical perspectives to strengthen conceptual understanding in secondary mathematics. Future studies employing controlled experimental designs are recommended to further validate and extend these findings.