Scaffolding the Scaffolder: Augmenting Parental Agency in the Developmental Arena with Generative AI
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Parent-child scaffolding interactions are fundamental to child development, yet their fleeting and complex nature makes it difficult for caregivers to self-monitor and enhance. We present GABRIEL, a novel artificial intelligence (AI) pipeline designed with generative AI (GenAI) as a reflective mirror to augment caregiver’s cognitive agency for children in dialogical scaffolding early reading interactions. The iterative process of GABRIEL’s development, detailing how established psychological frameworks, like an adult sensitivity scale and dialogical book sharing intersubjectivity, were operationalized to analyze parent-child reading not as an evaluative tool, but for using the reflection the GenAI provide. Quantitative analysis demonstrated the system’s reliability against a human-coded benchmark, establishing its viability as a trustworthy monitoring tool. This study argues for a paradigm shift in educational technology: moving from AI as a judge toward GenAI as a tool for extension that promotes the caregiver’s reflective effort. By providing a theory-informed mirror that challenges and engages the user, GABRIEL represents a promising approach for ethically integrating technology into the human arena of development and technologies.
