Exploring a Design Framework for Children's Agency through Participatory Design

Abstract

Children's agency plays a critical role in shaping children's autonomy, participation, and well-being in their interactions with digital systems, particularly in emerging child-AI contexts. However, how designers currently understand and reason about children's agency in practice remains underexplored. In this paper, we examine designers's engagement with children's agency through a participatory workshop in which we introduce a design-for-agency framework that supports designers externalising the consideration of agency in their design contexts. We find that while participants are committed to implementing ethical AI systems for children, they often struggle to understand why agency matters and how it can be operationalised in practice. Our agency design framework provided designers with a structured way to translate implicit, experience-based judgments into explicit articulation of agency trade-offs while acknowledging the associated design complexity. We conclude by offering initial insights into supporting designers' reasoning about children's agency and outlining directions for future research.

Figures (5)

• Figure 1: The four-stage reasoning process of the CHAI framework, showing how designers’ reasoning about children’s agency is transformed from an initial design context into explicit agency design through assessment, mapping, application, and reflection. The process produces intermediate reasoning artefacts and allows iteration among later stages.
• Figure 2: Four types of agency used in the CHAI framework as reasoning lenses. The visual abstractions represent different loci of action, control, and coordination across key stakeholders in each agency type. This illustration is used as part of the workshop materials to support designers’ reasoning about types of agency during the design process.
• Figure 3: Three levels of agency used in the CHAI framework as reasoning lenses. The levels capture differences in children’s involvement in initiating actions, influencing decisions, and shaping outcomes, as well as how these processes are mediated by other stakeholders and the AI system. This figure is used as part of the workshop materials in this study.
• Figure 4: Agency mapping matrix used in the CHAI participatory design workshop. Designers place key system functions along the top row and annotate each function with intended levels of support (low, medium, high) across four types of agency: individual, co-, proxy, and collective.
• Figure 5: Pre- and post-workshop self-reported agency understanding. Q1 evaluates participants' conceptual understanding of agency. Q2 understands stakeholder awareness. Q3--Q6 estimate concept translation to design.