<![CDATA[Background: The objective of this study was to analyze Worked Example Fading (WEF) implementation and evaluate its effectiveness in reducing instructional dependency while enhancing mathematical problem-solving abilities among fifth-grade elementary students. Method: This Classroom Action Research (CAR), utilizing the Kemmis and McTaggart model, spanned two cycles at SD Muhammadiyah Kabupaten Batang, Indonesia. Data were gathered via classroom observations, field notes, and Polya-based problem-solving tests. Crucially, the newly developed Worked Example Dependency Assessment (WEDA) monitored real-time trajectories of cognitive independence across five core dimensions: initiating problem solving, procedural independence, transfer ability, mathematical reasoning, and learning persistence. Result: The findings indicated that: (1) Baseline data showed acute dependency with 31.25% classical mastery. (2) Cycle I abrupt fading caused cognitive bottlenecks, though mastery reached 56.25%. (3) Cycle II adaptive micro-fading with self-explanation prompts minimized extraneous cognitive load, achieving an 82.10 average and 87.50% mastery. (4) WEDA evaluations unraveled a systematic mechanism of change, proving that procedural independence liberated working memory capacity for higher-order mathematical reasoning and far-transfer tasks. Implication: Elementary teachers should transition from rigid direct instruction to dynamic micro-fading strategies that shift cognitive agency to students. Furthermore, policymakers should champion process-oriented formative frameworks over purely summative assessments. Novelty: This study uncovers the dynamic, longitudinal cognitive transition from dependence to autonomy by operationalizing the novel WEDA process-based assessment framework in an authentic classroom]]>
