What is it about?
How can we tell if students are learning while playing an educational video game without giving them a formal test? In this study, we tracked how over 600 middle schoolers played a 3D science game. We used a smart computer model to connect their in-game actions, like how they used tools or built arguments, to their actual learning gains. We found that specific gameplay patterns are strong predictors of learning success, which helps us design more effective and adaptive educational games.
Featured Image
Photo by orva studio on Unsplash
Why is it important?
Traditional tests can interrupt the flow of learning and cause anxiety. This research is important because it provides a data-driven method to measure learning "invisibly" by analyzing how students naturally play an educational game. Our findings show that how a student plays is a powerful and reliable predictor of what they learn, paving the way for assessment to become a seamless and engaging part of the learning process itself.
Perspectives
For my co-authors and me, this project was driven by a simple question: What if we could understand what students are learning without ever making them feel like they're being tested? We've always been passionate about the power of games to create immersive worlds, and we believe we can make learning just as engaging. Our hope is that this work shows how we can use technology not just to deliver information, but to create more natural and joyful learning experiences where curiosity is the main driver, not the pressure of a final exam.
Wenyi Lu
University of Missouri Columbia
Read the Original
This page is a summary of: From In-Game Behaviors to Learning Gains: Constructing Bayesian Networks for Stealth Assessment, Proceedings of the ACM on Human-Computer Interaction, October 2025, ACM (Association for Computing Machinery),
DOI: 10.1145/3748615.
You can read the full text:
Contributors
The following have contributed to this page
