Collaborative Research: AI-Enhanced Problem-Posing Based Learning:Supporting Mathematics Teachers' in-the-Moment Instructional decision Making

This project will build and study an AI tool that supports teachers to make effective instructional decisions in problem-posing based learning (P-PBL). It provides students with opportunities to engage in authentic mathematical practices. The project seeks to develop an AI-integrated problem-posing system that is available for teachers and scholars. This project will increase understanding of how AI can work alongside teachers to support—not replace—their professional judgment when making instructional decisions in real time. The knowledge gained from this project will help guide the future development of AI tools that support teaching and learning across STEM fields. The project also has potential to help more students develop the mathematical knowledge, problem-solving skills, problem-posing skills, and confidence needed for success in college, careers, and an increasingly technology-driven world. Building on the work of two prior NSF-funded projects, this multi-method project will use AI to interpret students’ mathematical thinking based on extremely open tasks (i.e., problem-posing tasks) and partner with teachers to understand how best to support their instructional decision making with AI-generated information. The specific project goals are to: (1) develop an AI-based automatic assessment system that classifies student-posed problems based on learning goals and pedagogically useful parameters; (2) develop automatically generated reports (AutoRs) that help teachers analyze, select, sequence, and connect their students’ posed problems; (3) examine how AutoRs support teachers' in-the-moment instructional decision-making that promotes ambitious teaching; and (4) examine how AI-enhanced P-PBL can support cognitive and affective aspects of students' learning. Results from this work will inform the continuing discussion on the use of AI to analyze students' responses to extremely open tasks. It will also produce knowledge about how to support teachers’ instructional decision-making using AI-generated information. As AI tools to assist teachers continue to proliferate, their design depends on information critical to supporting teachers’ responsive instructional decision-making. To respond to this challenge, results from this project will help educators better understand and meet these challenges by identifying information that teachers need when handling students’ posed problems, in particular information from an AI system, and formulating a theory of teachers' in-the-moment decision-making based on AutoRs. 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.