Skip to main content Skip to navigation

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

Let us know you agree to cookies