This study analyzes the design and operational outcomes of a “PBL-AI Fusion Instructional Model” to bridge competency gaps and prevent “cognitive dependency” in basic SW liberal arts classes. Redefining GenAI as an “intelligent learning assistant,” the model employs a FILA-based “Task Activity Sheet” as core pedagogical scaffolding to ensure autonomous control over technology. The model is predicated on four principles: Contextual Reality, Cognitive Scaffolding, Reflective Subjectivity, and Integrated Flexibility. It utilizes a five-step metacognitive process—Planning, Check 1, Action, Check 2, and Evaluation—enabling learners to diagnose knowledge gaps and independently verify AI-suggested solutions. Analysis of a 15-week course at J University revealed significantly alleviated coding barriers and improved software self-efficacy. Notably, the “inexperienced learners” group outpaced “experienced learners” in control structure utilization (CT13) and flow analysis (CT10), effectively narrowing the competency gap regardless of prior programming experience. Qualitative results confirmed the formation of “Critical AI Literacy,” as students selectively adopted AI code, such as try-except syntax or Flag variables, based on their learning stage. Quantitatively, 96.7% of students found AI feedback useful, and 95.4% expressed intentions for continued use, indicating high satisfaction. The flow of thought recorded in activity sheets served as a vital data asset for personalized feedback. This research suggests that instructors must shift from “knowledge transmitters” to “learning facilitators.” Ultimately, this reflection-centered AI collaboration model offers a solution for driving qualitative innovation and fostering talents capable of autonomously managing and controlling AI technology across various disciplines.
Eun-Sill Jang (Thu,) studied this question.