PedaCo-Gen: Scaffolding Pedagogical Agency in Human-AI Collaborative Video Authoring
Injun Baek, Yearim Kim, Nojun Kwak
TL;DR
PedaCo-Gen tackles the misalignment between visually compelling AI-generated instructional videos and instructional quality by embedding Mayer's CTML into an interactive, three-phase authoring workflow that uses an intermediate representation as a reviewable blueprint. An expert LLM-based reviewer provides CTML-guided feedback, reframing AI output as a metacognitive scaffold that preserves educator agency. In a study with 23 educators across three topics, the system significantly improves CTML-based video quality and production efficiency compared with Baseline, with robust gains across all 12 CTML principles and positive but variable intentions to apply in practice. The work demonstrates that principled human-AI co-creation can enhance instructional design, trust, and efficiency, and outlines pathways for adaptive content, XAI, and broader curricular alignment in future AI authoring tools.
Abstract
While advancements in Text-to-Video (T2V) generative AI offer a promising path toward democratizing content creation, current models are often optimized for visual fidelity rather than instructional efficacy. This study introduces PedaCo-Gen, a pedagogically-informed human-AI collaborative video generating system for authoring instructional videos based on Mayer's Cognitive Theory of Multimedia Learning (CTML). Moving away from traditional "one-shot" generation, PedaCo-Gen introduces an Intermediate Representation (IR) phase, enabling educators to interactively review and refine video blueprints-comprising scripts and visual descriptions-with an AI reviewer. Our study with 23 education experts demonstrates that PedaCo-Gen significantly enhances video quality across various topics and CTML principles compared to baselines. Participants perceived the AI-driven guidance not merely as a set of instructions but as a metacognitive scaffold that augmented their instructional design expertise, reporting high production efficiency (M=4.26) and guide validity (M=4.04). These findings highlight the importance of reclaiming pedagogical agency through principled co-creation, providing a foundation for future AI authoring tools that harmonize generative power with human professional expertise.