Enhancing Chinese Children’s Narrative Skills Through AI-Supported Storytelling: A Comparative Study of Scaffolding Strategies and Transfer Effects

• In Event: English Language Learners and Literacy Proficiencies

Abstract

This study examines the impact of three AI-supported storytelling methods on Chinese children’s narrative performance, guided by the ICAP framework. 180 children aged 5 to 8 years were assigned to one of three conditions using an AI voice assistance with a generative model: AI-Guided Story Modeling, Autonomous Storytelling with AI Feedback, and Iterative Storytelling with AI Feedback. Narrative performance was assessed using the Mandarin version of MAIN across baseline, scaffolding, and transfer assessments. Results showed that the Iterative Storytelling condition significantly improved narrative skills, particularly in story structure and internal state terms. However, these effects disappeared when controlling for child-AI interaction duration. The study provides evidence for the AI systems in enhancing children’s narrative skills within a Chinese cultural context.

Authors

• Na Zhang, Capital Normal University

• Jinghan Xu, Capital Normal University

• Xianan Lu, Capital Normal University

• Chengcheng Wang, Capital Normal University

• Yifang Wang, Capital Normal University