การศึกษาผลกระทบของรูปแบบการเรียนรู้ AIM-4R ต่อการยกระดับความฉลาดรู้ด้านปัญญาประดิษฐ์

Abstract

This classroom action research investigates the enhancement of cognitive architectures in engineering education through the development, implementation, and empirical validation of the AIM-4R (Analyze, Implement, Monitor - Reporting, Relating, Reasoning, Reconstructing) micro-learning framework. Positioned within the context of generative artificial intelligence proliferation, this study addresses the critical challenge of cognitive over-reliance and automation bias among engineering students. The sample comprised 42 third-year manufacturing engineering students at Suranaree University of Technology (SUT) enrolled in the Industrial Database Management course (ENG35 4228), selected via purposive sampling. Utilizing a quasi-experimental one-group pretest-posttest design, the study deployed an array of instruments, including four structured micro-modules, an industrial performance-based task, a validated 20-item AI Literacy Scale, and a qualitative reflection matrix.

Quantitative findings revealed a statistically significant escalation in comprehensive AI literacy, with overall scores rising from a pre-test baseline (x̄ = 2.78, S.D. = 0.54$) to a post-test peak (x̄ = 4.52, S.D. = 0.38, t (41) = 18.42, p < .05$). Notably, the critical evaluation dimension exhibited the most profound expansion. Performance-based metrics demonstrated a substantial behavioral shift; students capable of executing context-rich prompt engineering and rigorous technical verification advanced from 32% to 89%, successfully detecting complex logical hallucinations embedded within database schemas. Qualitative content analysis of the 4R reflection sheets indicated a cognitive transition from descriptive reporting to high-order socio-technical reconstruction, wherein students formulated standardized manual cross-verification protocols using industrial engineering benchmarks. The study concludes that the AIM-4R framework serves as a robust pedagogical mechanism to transition engineering students from passive technology consumers to vigilant cognitive supervisors, establishing a baseline for algorithmic accountability in the Industry 5.0 paradigm.