ChatGPT-Assisted Lesson Planning for Children’s STEAM Arts Education: An Experimental Study on Benefits, Challenges, Methods, and a Prompt Framework

This study focuses on art-led STEAM education to cultivate ‘21st-century 4C skills’ (communication, collaboration, critical thinking, and creativity). It explores the potential of ChatGPT in enhancing lesson planning within STEAM art education. An experimental study was conducted comparing traditional lesson plans with ChatGPT-assisted plans, involving 13 teachers and 6 professors. A mixed-methods approach was employed: qualitative data (semi-structured interviews, ChatGPT interaction records, etc.) were analyzed thematically, while quantitative data (questionnaire surveys) were examined using descriptive and inferential statistical analyses. Results demonstrate that ChatGPT significantly improved the quality of lesson planning, with ChatGPT-assisted plans achieving an average score of 21.0 compared to 17.0 for traditional plans. In addition, ChatGPT also offers benefits such as enhanced efficiency and supported interdisciplinary and creative approaches. However, challenges remain regarding content personalization, usability, and logical reasoning in image generation. Beyond these findings, the study proposes key methods for AI-assisted lesson planning, highlighting steps in generating lesson plans and searching for multimedia resources. In addition, a prompt framework is also developed to guide the generation process, including structured initial prompts and iterative optimization based on evaluation dimensions. Overall, the results provide both theoretical and practical insights into the application of generative AI in education.