<![CDATA[Indonesia’s Kurikulum Merdeka (Merdeka Curriculum) demands personalized mathematics learning, but teachers in heterogeneous junior high school classrooms lack scalable tools. This creates a critical “implementation gap” between the policy’s differentiation ideals and classroom reality. This study aimed to design, develop, and evaluate a bespoke AI-Based Adaptive Learning System (ALS) natively aligned with the Merdeka Curriculum’s “Fase D” framework to enhance student achievement. A mixed-methods, quasi-experimental pre-test/post-test control group design (N=435) was employed over 14 weeks. The treatment group (n=232) used the ALS. Effectiveness was measured by an ANCOVA on mathematics achievement scores, supplemented by usability surveys (SUS) and qualitative data from teachers and students. The ALS group demonstrated significantly superior mathematics achievement (F(1, 432) = 121.4, p < .001, \eta_p^2 = .219). Qualitative data confirmed the system enabled “Targeted Differentiation” for teachers and provided “Personalized Pacing and Safety” for students, reducing peer anxiety. The context-aligned ALS is a highly effective, usable (SUS=79.5), and feasible tool. It successfully provides the scalable, individualized support required to operationalize the pedagogical philosophy of the Kurikulum Merdeka.]]>
Copyrights © 2025
