GENERATIVE AI INSIDE THE EDUCATIONAL DESIGN LIFE CYCLE: A CASE STUDY IN PEAK TEACHING

Abstract

A foundational principle of student-centered learning is that effective instruction must meet students where they are. Yet delivering truly differentiated instruction at scale has remained elusive—the logistical burden of creating multiple learning pathways often exceeds what a single instructor can manage. This study examines whether AI-assisted course design can make differentiated pathways practical and effective in an introductory Business Law course. Results demonstrate that scaffolded pathways significantly improved learning gains for underprepared students (β = 23.69, p < .001), while a gamified alternative produced a modest benefit compared to historical instruction (d = 0.26). Critically, sentiment analysis of student feedback revealed friction points that informed targeted improvements, demonstrating a complete action research cycle enabled by AI tools at each stage.