Reconstructing engineering-oriented STEM curriculum through an evidence-driven recursive learning model: insights from deep-tech product development
摘要
Engineering-oriented STEM curriculum at the upper secondary level aims to prepare students to engage with complex technological challenges. However, classroom implementations frequently reduce engineering practice to fragmented projects or activity-based tasks, failing to reflect the structural logic and epistemic demands of authentic engineering work—particularly in deep-tech product development contexts. Drawing on complex adaptive systems and situated learning theories, this study conducted a longitudinal case study of a Chinese team developing autonomous drone-charging systems, analyzing engineers’ learning mechanisms across four stages of product development: needs definition, R&D design, testing and iteration, and delivery and improvement. Findings reveal engineering learning as an evidence-constrained recursive process. Across stages, learning shifts from exploratory clarification to data-driven optimization; decision-making evolves from intuition-informed judgment to explicit evidence-based reasoning; collaboration transforms from task coordination to evidence-mediated collective deliberation. Structurally, learning operates through a claim–evidence–decision–consolidation cycle. Based on these findings, this study proposes an evidence-driven recursive learning model and a four-stage curriculum design framework incorporating staged evidence thresholds. The framework offers a structural logic for organizing engineering-oriented STEM curriculum at the upper secondary level.