수업설계에 적용된 생성형 AI 프롬프트의 효과

초록

This study aimed to explore the application of six types of Generative AI prompts in the instructional design process—instructional analysis, performance objective, assessment instrument, and instructional strategy—and to conduct a comparative analysis of the resulting outputs. While no significant differences were observed among the prompts in terms of general target analysis, the Persona prompt demonstrated superior specificity and depth in sub-function analysis, and was therefore adopted as the benchmark for subsequent design stages to ensure coherence and consistency. In a further comparison between the Persona prompt and the Suggested prompt—which enables Generative AI to autonomously select the most contextually appropriate prompt type—the AI exhibited an adaptive selection pattern: Zero-shot and Few-shot were primarily chosen for lower-order cognitive tasks involving knowledge and comprehension; Role-playing and Persona were selected for tasks requiring application and analysis; and Chain of Thought (CoT) and Zero-shot with CoT were applied to tasks demanding higher-order cognitive engagement such as error correction or system construction. This indicates that the AI was capable of selecting prompts in a manner sensitive to both content characteristics and cognitive complexity. The validity of the AI-generated outputs was evaluated by experts in instructional systems design and content, with both content and format validity rated as satisfactory. Given that instructional design is an inherently complex and cognitively demanding process that necessitates iterative feedback and forward-thinking considerations, the findings of this study suggest that Generative AI holds significant potential to enhance the reliability, validity, and overall quality of instructional design practices. Accordingly, this research offers meaningful empirical insight into the practical integration of AI in educational development contexts.

키워드