Collaborative Research: Using Large Language Models to Provide More Dynamic and Effective Learning Support in a Digital Learning Game

Many students struggle with mathematics, and digital learning games have emerged as a promising way to increase engagement and improve learning in STEM education. However, the instructional support provided by most educational games remains limited, relying on static hints that cannot respond to individual students' needs. Recent advances in Artificial Intelligence (AI), particularly Large Language Models (LLMs), create new opportunities to provide more personalized assistance, but little is known about how these technologies can be used effectively and responsibly in learning environments. This project addresses this challenge by investigating whether AI can provide individualized support to elementary and middle school students learning decimal concepts through a digital learning game. By advancing understanding of how AI can support learning at scale, the project contributes to improving STEM education, expanding access to high-quality instructional support, and strengthening the nation's capacity to develop effective educational technologies. The project will investigate whether dynamically generated, LLM-based hints improve student learning, engagement, and productive learning behaviors compared to traditional static hints or no hints. Researchers will enhance Decimal Point, a research-based mathematics learning game, to generate personalized hints that respond to students' problem-solving histories, misconceptions, and learning behaviors. Approximately 800 fifth- and sixth-grade students will participate in classroom studies evaluating the effectiveness of these hints. The project will examine when, why, and for whom personalized hints are most beneficial, while also studying the practical challenges of implementing AI in educational technologies. Outcomes will include new knowledge about adaptive instructional support, design guidelines for educational game developers, publicly available software and research resources, and a freely available enhanced version of the Decimal Point learning game for classroom use. This project was funded by the STEM K-12 program, which supports fundamental, applied, and translational research on STEM learning in both formal K-12 education and informal learning environments. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.