AutoTutor 3-D Simulations: Analyzing Users’ Actions and Learning Trends

Research in simulations and their educational efficacy has had mixed success in the educational and cognitive sciences. There are challenges in understanding the complex nature of simulations, critical points of learner interactivity with computer simulations, and appropriate methods for testing and discovering potential benefits. The current research investigated learning from AutoTutor, an intelligent tutoring system, that interacts with learners in natural language and that launches embedded interactive 3-D simulations for tutoring conceptual physics. An experiment on college students shed light on conditions that promote learning, with results that hopefully will scale up in diverse educational settings with learning technologies.