Ë®¶à¶àµ¼º½

Imagine that a middle school student – at the click of a button – could use generative AI (GenAI) to produce a mathematically precise, beautifully rendered image to bring a difficult math problem to life.

Through a $899,988 grant from the National Science Foundation, Candace Walkington, Ph.D., and her research team will examine how AI-generated diagrams could improve math learning for middle school students. 

Walkington, the Annette and Harold Simmons Centennial Chair and Professor at the Simmons School of Education and Human Development, will serve as the Principal Investigator for the study. Her colleague at Simmons, Professor Jiun-Yu “Jerry” Wu, Ph.D., will serve as a Co-Principal Investigator (Co-PI). The project builds on Walkington’s past work with the Louisiana-based AI startup Ailectric, a collaborator on the grant, with lead engineer Dr. Robert DiBiano.

Walkington’s research proposes a novel approach to making middle school math instruction more relevant and accessible to students from all backgrounds. Math word problems are particularly challenging for many students. The high-quality illustrative diagrams, created through open-source GenAI image-generating models that use ControlNet functionality, help students better visualize the problem. 

“ControlNets allow for rich illustrative features to be layered over mathematical diagrams that the GenAI renders from natural language instructions, creating free, high-quality visuals to accompany a broad range of mathematical learning tasks,” Walkington said. 

Walkington will initially interview teachers and Open Educational Resource (OER) developers. Leading Open Education Resource researcher Virginia Clinton-Lisell, Ph.D., from the University of North Dakota, a Co-PI for the project, and her team will pilot their AI-generated illustrative diagrams and methods for developing those diagrams with K-12 and higher education teachers. Finally, they will conduct an experiment in the ASSISTments online homework platform to test the effect of AI-generated illustrative diagrams on middle school students’ mathematical reasoning. Cristina Heffernan with the ASSISTments Foundation is also a Co-PI on the project and works with Walkington’s team on additional AI projects.

Walkington says her team will seek to advance research on how high-quality OER can be delivered to students in a cost-effective way, so the resources can better compete with costly commercial curricula broadly impact a wide range of students.

“More research in AI in education is needed to determine how illustrative math visuals and open educational resources shape students’ mental representations and facilitate their interest and performance in mathematics learning,” Wu said. “It is exciting to know this research can help answer those questions.”

This material is based upon work supported by the National Science Foundation under Award Number (FAIN): 2507009.

Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.