<![CDATA[This study applies Educational Data Mining (EDM) to examine how students’ mathematical communication skills and cognitive styles interact under different instructional contexts. A quantitative research that applies quasi-experimental design combined with EDM was employed involving 64 seventh-grade students at SMP Negeri 2 Jaten, divided into Learning Cycle 7E (LC7e) and Direct Instruction (DI) groups. The research instruments included a validated essay test for mathematical communication and the Group Embedded Figures Test (GEFT). Data were analyzed using the Fuzzy C-Means (FCM) algorithm to partition students into distinct clusters based on their cognitive-communicative attributes. The analysis resulted in a stable five-cluster model, representing progressive levels of cognitive–communicative integration. The model shows that cognitive profile predominantly creates cluster structure, whereas the Learning Cycle 7E (LC7E) model exerts a moderating influence. Students taught through LC7E were more concentrated in higher-performing clusters than those in DI classrooms. Furthermore, Field-Independent (FI) learners tended to achieve the highest communicative profiles, yet Field-Dependent (FD) learners also benefited meaningfully from LC7E activities that emphasized exploration and reflection. These results demonstrate that the LC7E model supports cognitive and communicative development across the learner spectrum, with differentiated gains linked to cognitive style. These findings highlight the utility of EDM in capturing student heterogeneity and provide a basis for educators to design adaptive learning strategies that accommodate diverse cognitive characteristics. ]]>
Copyrights © 2026
