Fulfilling K-12 Students' Mathematics Needs: Near Peer Teaching and Scientific Intelligent Clustering Tutor Approach

Abstract

This investigation delves into the pressing need for curriculum reform in Indonesia, specifically examining the "Merdeka Belajar: Kampus Merdeka (MBKM)" policy, with a focus on the Near Peer Teaching (NPT) model. Previous studies have flagged NPT's shortcomings, attributing them to inconsistent tutor feedback rooted in relational challenges. Despite positive anecdotal evidence of student transformation through NPT, such accounts often lack objectivity. This research strategically surveys K-12 vocational schools in Kuningan, honing in on challenges in Mathematics to inform responsive teaching strategies. Noteworthy is the persistence of selecting peer tutors based on final exam scores, a practice upheld despite the initial randomness in NPT tutor selection, creating hurdles in gauging effectiveness. Paradoxically, empirical data suggests that third-semester students make better tutors, yet the fixation on final exam scores persists. To propel the NPT model forward, the study advocates for clustering tutors based on scientific intelligence, integrating the innovative application of machine learning algorithm K-Means. This comprehensive approach melds quantitative data science with qualitative Deep Interviews, aiming to refine and optimize the identification of suitable peer tutors. The crux of the findings revolves around the imperative to refine the selection process for peer tutors, considering factors such as interest, motivation, and academic achievement, to significantly amplify the efficacy of NPT within the learning environment..