Table of Contents
Fetching ...

"My lollipop dropped..."-Probing Design Opportunities for SEL Agents through Children's Peer Co-Creation of Social-Emotional Stories

Hanqing Zhou, Anastasia Nikolova, Pengcheng An

TL;DR

This study investigates how to design child-appropriate social-emotional learning (SEL) storytelling agents through a peer co-design process with preschoolers. The approach combines hands-on crafting and AI-assisted visualization via Midjourney across two 1.5-hour sessions with eight children, yielding four candidate agents and 14 co-created social-emotional events. Key findings identify desirable agent characteristics, diverse emotional events, and rich verbal/nonverbal dynamics that can inform agent behaviors and interactions. The work provides concrete design implications for multi-modal, proactive SEL agents and outlines limitations and directions for future research in HCI with young children.

Abstract

This Late-Breaking Work explores the significance of socio-emotional learning (SEL) and the challenges inherent in designing child-appropriate technologies, namely storytelling agents, to support SEL. We aim to probe their needs and preferences regarding agents for SEL by conducting co-design which involves children co-creating characters and social-emotional stories. We conducted collaborative story-making activities with children aged four to six years old. Our findings could inform the design of both verbal and nonverbal interactions of agents, which are to be aligned with children's understanding and interest. Based on the child-led peer co-design, our work enhances the understanding of SEL agent designs and behaviors tailored to children's socio-emotional needs, thereby offering practical implications for more effective SEL tools in future HCI research and practice.

"My lollipop dropped..."-Probing Design Opportunities for SEL Agents through Children's Peer Co-Creation of Social-Emotional Stories

TL;DR

This study investigates how to design child-appropriate social-emotional learning (SEL) storytelling agents through a peer co-design process with preschoolers. The approach combines hands-on crafting and AI-assisted visualization via Midjourney across two 1.5-hour sessions with eight children, yielding four candidate agents and 14 co-created social-emotional events. Key findings identify desirable agent characteristics, diverse emotional events, and rich verbal/nonverbal dynamics that can inform agent behaviors and interactions. The work provides concrete design implications for multi-modal, proactive SEL agents and outlines limitations and directions for future research in HCI with young children.

Abstract

This Late-Breaking Work explores the significance of socio-emotional learning (SEL) and the challenges inherent in designing child-appropriate technologies, namely storytelling agents, to support SEL. We aim to probe their needs and preferences regarding agents for SEL by conducting co-design which involves children co-creating characters and social-emotional stories. We conducted collaborative story-making activities with children aged four to six years old. Our findings could inform the design of both verbal and nonverbal interactions of agents, which are to be aligned with children's understanding and interest. Based on the child-led peer co-design, our work enhances the understanding of SEL agent designs and behaviors tailored to children's socio-emotional needs, thereby offering practical implications for more effective SEL tools in future HCI research and practice.
Paper Structure (13 sections, 3 figures, 2 tables)

This paper contains 13 sections, 3 figures, 2 tables.

Figures (3)

  • Figure 1: Illustration for child-AI co-creating: integrating physical materials and generative AI in the creative process.
  • Figure 2: The four characters selected by the children in Co-design Session 1
  • Figure 3: Guidelines and Examples on Midjourney Prompts.