Exploring the Impact of an LLM-Powered Teachable Agent on Learning Gains and Cognitive Load in Music Education
Project Overview
The document explores the transformative role of generative AI in education, focusing on the implementation of a large language model (LLM)-powered teachable agent named Chat Melody in music education. A study involving 28 Chinese university students revealed that those using Chat Melody experienced significant learning gains and a reduction in cognitive load compared to traditional self-directed learning approaches. By leveraging the Learning by Teaching (LBT) pedagogy, Chat Melody fosters structured dialogue and active engagement in music theory, enhancing the learning experience. The findings underscore the potential of generative AI technologies to create more effective and interactive educational environments, facilitating deeper understanding and retention of complex subjects through personalized and adaptive learning experiences. Overall, the integration of AI in educational settings demonstrates promising outcomes, paving the way for innovative teaching methods that can improve student engagement and mastery of content.
Key Applications
Chat Melody, an LLM-powered teachable agent
Context: Music theory learning for university students with prior instrumental experience
Implementation: Students engaged with the teachable agent in structured dialogues while analyzing music, as opposed to self-directed learning with instructional materials.
Outcomes: Students using Chat Melody achieved higher post-test scores and reported lower cognitive load.
Challenges: Limitations include a small sample size and the need for future studies to assess long-term retention.
Implementation Barriers
Technical Barrier
The integration of LLM-powered agents in traditional education systems is still underexplored, particularly in non-STEM disciplines like music.
Proposed Solutions: Further research and development of AI tools tailored for specific educational contexts, such as music education.
Project Team
Lingxi Jin
Researcher
Baicheng Lin
Researcher
Mengze Hong
Researcher
Kun Zhang
Researcher
Hyo-Jeong So
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Lingxi Jin, Baicheng Lin, Mengze Hong, Kun Zhang, Hyo-Jeong So
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai