Integrating Problem-Based Learning, Tri-N, and Deep Learning to Develop Critical Thinking in 3T Elementary Science Education

Abstract

Developing critical-thinking skills in disadvantaged and geographically constrained elementary schools remains challenging because conventional instruction offers limited opportunities for inquiry, contextual learning, and higher-order thinking. This study examined the implementation and classroom-level outcomes of an integrated Problem-Based Learning (PBL), Tri-N (Niteni–Nirokke–Nambahi), and Deep Learning framework for developing fifth-grade students’ critical-thinking skills in a 3T (frontier, outermost, and least-developed) elementary school in Kupang Regency, Indonesia. A classroom action research design within a convergent mixed-methods framework involved 28 students across two four-week cycles. Quantitative data were obtained from critical-thinking assessments, while qualitative evidence came from observations, student work, interviews, and teacher reflections. Mean critical-thinking scores increased from 57.5 at baseline to 68.2 in Cycle I and 78.6 in Cycle II, while mastery increased from 32.1% to 60.7% and 85.7%, respectively. The largest gains occurred in evaluation and inference. Improvements were also observed in environmental observation, explanatory questioning, integration of local cultural knowledge, and confidence in developing contextual solutions. Implementation fidelity increased from 78.5% to 92.3% following contextual refinements. The findings indicate that the integrated framework offers a context-responsive pathway for strengthening critical-thinking practices in resource-constrained 3T education, while causal claims remain limited by the single-school design and absence of a comparison group.