Who Tells the Better Math Story? Exploring Gen AI in a Digital Algebra Game

Abstract

Learning algebra can be challenging for children due to its abstract nature. Story-based game design can support learning by contextualising mathematical structures. Recent advances in generative artificial intelligence (genAI) offer new opportunities for reflection and meaning-making. This paper presents a genAI-supported, story-based digital algebra game designed for students aged 9–11 encountering algebra for the first time. The game consists of three levels, progressing from recognition to production.. An embedded LLM system (Algy) generates an alternative story using the same student inputs - character, place, and equation - for creating their own stories. Students compare both stories, select one, and justify their choice. The study involved 136 students and four teachers. Data from written justifications and interviews were analysed thematically. Results showed that students choosing AI-generated stories valued clarity, detail, correctness, and perceived intelligence. In contrast, those preferring their own stories emphasised ownership, creativity, and emotional engagement. Teachers reported that integrating genAI within a structured game made it more controllable and trustworthy. Findings suggest that structured integration of genAI can support reflection while preserving student ownership.