Intelligent tutoring systems are adaptive learning environments designed to support individualized instruction. The adaptation embedded within these systems is often guided by user models that represent one or more aspects of students' domain knowledge, actions, or performance. The proposed project focuses on the development and testing of user models within the iSTART-2 intelligent tutoring system, which will be informed by dynamic methodologies and data mining techniques. My previous work has used post hoc dynamic methodologies to quantify optimal and in-optimal learning behaviors within the game-based system, iSTART-2. I plan to build upon this work by conducting dynamical analyses in real-time to inform the user models embedded within iSTART-2. I am seeking advice and feedback on the statistical methods and feature selection that should included within the new dynamic user model. The implications of this approach for both iSTART-2 and the EDM field are discussed. [For complete proceedings, see ED560503.]