<![CDATA[Purpose of the study: Despite widespread recognition of Realistic Mathematics Education (RME) and student learning interest as critical factors in mathematics learning, their simultaneous contribution to performance, as perceived by classroom teachers, remains insufficiently examined. This study investigates the concurrent relationships between RME implementation quality and student learning interest with teacher-rated mathematics learning performance in elementary schools. Methodology: A quantitative cross-sectional survey involved n = 55 purposively selected elementary school mathematics teachers. Using validated questionnaires with 5-point Likert scales, teachers rated RME practices, student interest, and mathematics performance across nine competency dimensions based on year-long classroom observations. Structural Equation Modeling–Partial Least Squares (SEM-PLS) with bootstrapping (5,000 resamples) was employed. The measurement model demonstrated robust psychometric properties (AVE > 0.70, CR > 0.93, HTMT < 0.90). Main Findings: RME implementation significantly predicted teacher-rated performance (β = 0.468, p < 0.001, f² = 0.312, medium-to-large effect), as did student learning interest (β = 0.382, p < 0.001, f² = 0.198, medium effect). Both predictors jointly explained 58.4% of performance variance (R² = 0.584, Q² = 0.398), indicating substantial explanatory and predictive capacity. Novelty/Originality of this study: This study uniquely integrates pedagogical and psychological predictors within a single SEM-PLS framework from the practitioner perspective, addressing a methodological gap in Indonesian elementary mathematics research that has predominantly relied on separate, small-scale analyses. Findings carry direct implications for teacher professional development and curriculum design, particularly in advancing RME adoption and interest-fostering strategies within Indonesia’s Merdeka Curriculum context.]]>
