Enhanced Learning Through the Integration of AI and Best Practices: A Randomized Controlled Study Comparing AI-Supported Pedagogy with Active Learning in Introductory Physics

ORAL  · Invited

Abstract

We present a randomized controlled study that compares learning gains and perceptions of AI-supported lessons in an introductory physics course to lessons that employ active learning with instructor and peer support. This study implements a generalizable approach to AI-supported instruction that offers a potential solution to long-standing challenges in STEM education. We find that introducing concepts to students using an AI-supported environment can lead to improved engagement, motivation, quicker learning and learning gains approaching a standard deviation.

Presenters

  • Kelly Miller

    • Harvard

Authors

  • Kelly Miller

    • Harvard
  • Greg Kestin

    • Harvard