This project explores how students learn STEM content in educational games and how their learning can be assessed and analyzed using biopsychological sensor and game event data streams. Biopsychological sensors including eyetrackers, electrodermal response sensors, heart rate sensors, etc. have long been used in the field of psychology to improve our understanding of memory, cognition, affect, and learning. Prior research has focused primarily on tightly controlled lab studies with limited ecological validity. This project aims to apply those measures in a novel way to study all elements of the cognitive learning process in a real-world educational game.