Comparing the Effectiveness of AI-Powered Scaffolding and Human Teacher Scaffolding on Instructional Design (Poster 34): SIG-Design and Technology, Stage 2, 8:12 AM

• In Event: e-Lightening Ed-Talk Session 18 - Stage 2

Abstract

To address limitations of general AI tools, we developed a 5E-model based chatbot to scaffold student instructional design. The study found the AI chatbot, combined with human teacher guidance, significantly improved student performance compared to teacher-only scaffolding. The chatbot provided valuable ‘explore’ phase suggestions on frameworks, activities, and topics. In ‘evaluate’, it offered feedback and alternative approaches for improvement. However, students felt the chatbot’s evaluation quality needed better alignment with assessment criteria. We recommend using AI chatbots to support instructional design conceptualization, while emphasizing teachers’ critical role in evaluating final designs and providing personalized support.

Authors

• Shurui Bai, The Education University of Hong Kong

• Donn Emmanuel Gonda, The University of Hong Kong