Quantum information science (QIS), which uses the laws of quantum physics to process and store information, is expected to broadly impact society through new developments in commerce, governance, privacy, employment, education, and other areas. Unfortunately, QIS is a challenging, interdisciplinary field to learn. The goal of this project is to advance QIS education by using virtual reality (VR) and machine learning to adaptively address misconceptions about the field. This project will leverage QubitVR, a VR application previously developed for learning foundational QIS concepts like superposition, measurement, and entanglement. The project will involve the following aims: identify and predict QIS misconceptions, develop and validate a new QIS Concept Introductory Test (QISCIT) for assessing learning outcomes, develop and compare two intelligent tutoring versions of QubitVR (one that employs proactive conceptual scaffolds based on the machine learning models and one that employs reactive scaffolds based on conventional action-condition rules-based reasoning), ecologically validate the efficacy of QubitVR in a longitudinal study, and develop desktop and smartphone versions of QubitVR, which will be made openly available alongside the VR versions for broader educational impacts.