<![CDATA[Developing junior high school students’ conceptual understanding through coherent mathematical representations remains a persistent challenge, as conventional assessments tend to prioritize procedural accuracy while neglecting how students construct, connect, and integrate symbolic, visual, and verbal forms. Differentiated learning is designed to address learner diversity, yet its role in fostering representational competence and providing diagnostic insight is still insufficiently examined. This study evaluates students’ mathematical representation skills within differentiated learning environments by identifying representational patterns, variations across readiness levels, and the diagnostic value of representation-focused assessment. Using a qualitative approach, data were collected through analysis of students’ written work and semi-structured interviews in classrooms implementing differentiated mathematics tasks. Participants were categorized into high, moderate, and low representational readiness levels. The findings reveal substantial variability in representational quality and integration. High-readiness students demonstrated flexible coordination across multiple representations, while moderate- and low-readiness students tended to rely on single or fragmented forms. Although differentiated tasks increased engagement and learner choice, effective representational coherence required explicit and targeted scaffolding. Importantly, representation-focused assessment uncovered learning gaps, misconceptions, and conceptual weaknesses that were often overlooked by procedural evaluations. The study concludes that differentiated learning can support the development of representational competence, but its effectiveness depends on intentional scaffolding and diagnostic assessment practices that promote deeper, more equitable conceptual understanding in mathematics.]]>
