From Scaffolding to Transfer: The Impact of AI-Powered Narrative Support on Children’s Narrative Skills

Published: 6 May 2025| Version 1 | DOI: 10.17632/ptsprmjdb5.1
Contributors:
Na Zhang, Jinghan Xu, Ran Wei, Yifang Wang

Description

This dataset contains narrative performance and interaction data from a study examining the impact of AI-powered narrative support on children’s storytelling skills. Below are detailed descriptions of all variables: 1.Participant Information Child_ID: Unique identifier. Age_months: Child’s age in months. Condition: Experimental group (Passive-Active Retelling, Iterative Storytelling with AI Feedback, Autonomous Storytelling with AI Feedback). Gender: Child’s gender (boy, girl). Child_AI_Interaction_Duration: Interaction time (minutes). 2. Narrative Scores Narrative skills were assessed for three stories (Cat, Dog, Baby Bird) across three dimensions: story structure, story complexity, and internal state terms. Scores are labeled by story (Cat, Dog, Baby Bird). [Story] A1–A20: Story structure subitem scores (e.g., CatA1, DogA1). [Story] A Total: Total core of story structure. [Story]B1–B4: Story complexity classification counts (B1=Attempt-Outcome, B2=Goal, B3=Goal-Attempt/Outcome, B4=Goal-Attempt-Outcome). [Story]B Total: Total score of weighted story complexity (=B1*1 + B2*2 + B3*3 + B4*4). [Story]C Total: Total score of internal state terms. 3.Subjective Ratings Enjoyment: Perceived enjoyment. Interest: Perceived interest.

Files

Categories

Digital Storytelling, Literacy, Quantitative Research in Education, Information Technology in Education

Funders

  • National Education Sciences Planning Project of China
    Grant ID: BHA230141

Licence