<![CDATA[Purpose:This study investigates the impact of the Connected Mathematics Project (CMP) model on eighth-grade students’ ability to form mathematical connections at SMP Negeri 1 Martapura. It also examines how varying levels of self-efficacy relate to these connection skills and whether an interaction exists between instructional model (CMP vs. discovery learning) and self-efficacy in influencing outcomes. Method:A quasi-experimental, posttest-only control-group design was used. Two groups of eighth graders were assigned: one received CMP-based instruction and the other discovery-learning lessons. After the intervention, all students completed a standardized connection-skills posttest and a validated self-efficacy questionnaire. Participants were categorized into high- and low-efficacy subgroups. Two-way ANOVA assessed the main effects of instructional model and self-efficacy, as well as their interaction. Findings:Students taught via the CMP model scored significantly higher on mathematical connection tasks than those in the discovery-learning group. Independently, students with high self-efficacy outperformed those with lower self-efficacy. No significant interaction effect was detected, indicating CMP’s benefit across confidence levels. Significance:Findings demonstrate that CMP is a robust approach for enhancing mathematical connection skills regardless of student self-efficacy. They underscore the value of building learner confidence alongside innovative pedagogy and support broader adoption of CMP to strengthen real-world mathematical understanding.]]>
