M-Powering Teachers: A Machine Learning Tool for Instructional Management and Feedback

On November 10th, TUM’s Center for Educational Technologies will host Prof. Heather C. Hill, Harvard University, who will present their platform M-Powering Teachers.

Abstract

This talk describes M-Powering Teachers (MPT), a machine-learning-based platform that automatically measures key mathematics instructional practices linked to student outcomes—including teacher focusing questions, student mathematical reasoning, and student and teacher mathematical vocabulary use—then provides visualized feedback through a user-friendly platform. Results from randomized controlled trials demonstrate promising impacts, with instructors showing significant improvements in targeted teaching practices. Current efforts focus on integrating this automated feedback into instructional coaching workflows to combine the efficiency of AI with the personalized benefits of human coaching. The project represents a significant advancement toward scalable, data-driven approaches to improving mathematics instruction in ways that support more equitable and effective teaching.

Guest Biography

Heather C. Hill studies policies and programs designed to improve teacher and teaching quality, particularly in mathematics. Her recent research focuses on teacher learning and professional development, the quality of mathematics instruction, the effectiveness of different approaches to teacher education, and machine learning tools that automate the measurement of instruction and feedback to teachers. Hill and her team have also created assessments that capture teachers’ mathematical knowledge for teaching and teachers’ mathematical quality of instruction and disseminated these assessments via online training and administration systems. Hill is an elected member of the National Academy of Education and the American Academy of Arts and Sciences.