Prompt Engineering as Pedagogical Interface: A Computational Study of Generative AI Systems for Enhancing Creative Thinking in Teacher Education

This study presents a computational framework investigating how generative AI systems enhance creative thinking through prompt engineering mechanisms in teacher education. A 12-week quasi-experimental intervention with 240 pre-service teachers employed natural language processing-based instructional design tools to analyze human-AI interaction patterns. Quantitative results demonstrate that AI-assisted computational workflows significantly enhance all six creative thinking dimensions: fluency, flexibility, originality, elaboration, overall creativity, and instructional design quality (Cohen's d=0.94-1.28, p0.75) yielding 15-20% higher creative outcomes through optimized human-AI information exchange. Computational path analysis establishes and empirically validates a tripartite mechanism model with demonstrated effectiveness: (1) cognitive load reduction pathway via AI-automated routine tasks showing 38% reduction in instructional design processing time with corresponding increases in creative output, (2) metacognitive development pathway through iterative prompt reflection cycles showing positive correlation with originality scores (r=0.64, p