Pedagogy-guided AI generates better STEM learning videos
From Content Generation to Learning Support: Pedagogy-Guided Generative Video Tutors for STEM Learning
Computation and LanguageArtificial IntelligenceComputers and Society
Summary
Creating educational videos with AI often overlooks how students learn best. The authors developed PIVOT, a system that uses teaching methods to guide the whole video creation process, ensuring videos are clear and helpful. Their videos also include ways to check if students understand the material and fix mistakes they might have. Tests showed these videos are well-organized and good for learning STEM subjects.
What this means in practice
- •For educational content creators: Create STEM learning videos with integrated teaching principles and assessment features to better support student understanding.
- •For corporate training teams: Produce technical instructional videos that include checks for learner misconceptions and personalized remediation.
- •For corporate elearning developers: Develop selling video tutoring platforms with pedagogy-guided generation and automated assessment to enhance employee skill training.$Commercial implications: Enables commercial video tutoring platforms that combine AI-generated content with verified pedagogy and learner feedback.
Authors
Xinchen Ma, Shuimu Wang, Gaole He, Yanbin Zhang, Chunyang Wang, Yunshi Lan, Weining Qian
Abstract
Generative AI enables scalable production of educational videos, but current systems largely focus on producing visually coherent content rather than supporting learning. As a result, generated videos often lack explicit pedagogical structure, reliable quality control, and mechanisms for assessing learner understanding or addressing misconceptions. In this work, we introduce PIVOT (Pedagogy-guided Instructional VideO Tutoring), a generative video tutoring framework for STEM learning via learning-centered instructional support.1 Inspired by conventional teaching workflows, our framework integrates pedagogy into the full generation pipeline: it first uses instructional principles to guide storyboard generation, then produces verified multimodal videos through code-centric generation and a pedagogical verification harness, and finally connects videos with assessment and misconception-aware remediation. Experiments and expert evaluations across four STEM domains show that our framework produces educational videos with pedagogically aligned content, clear and engaging presentation, coherent instructional flow, and perceived effectiveness for learning. These findings suggest a human-centered perspective on educational content generation: generative systems should be evaluated and designed not only by what they produce, but also by how they support teaching practices, learner understanding, and corrective feedback.