<![CDATA[The quality of learning in elementary education increasingly demands instructional approaches that strengthen students’ conceptual understanding and critical thinking. Preliminary observations at SD Negeri 79/IV Kota Jambi showed low performance, with conceptual understanding at 43.45% and critical thinking at 40.18%. This study implemented Deep Learning as a pedagogical approach (not an artificial intelligence/machine learning technique) through staged activities of interpretation, analysis, synthesis, and reflection in the Grade 5 IPAS ecosystem unit. Using a two-cycle Classroom Action Research design (planning, action, observation, and reflection), the intervention was conducted in regular lessons (70 minutes per meeting) with 28 students. Data were collected using concept understanding tests, critical thinking assessments, and structured observation sheets of teacher and student activities. Quantitative data were analyzed descriptively, while observation data were coded to capture engagement and implementation patterns. Results showed substantial gains across cycles: conceptual understanding increased from 43.45% (pre-action) to 68.20% (Cycle I) and 86.50% (Cycle II), and critical thinking improved from 40.18% to 65.10% and 84.75%. Teacher activity rose from 78% to 92% and student activity from 74% to 90%, indicating improved implementation and participation. These findings suggest that staged Deep Learning routines refined through CAR can enhance meaning construction and higher-order thinking in primary science learning.]]>
Copyrights © 2026
