Cognitive and learning theories that support Aptitude Treatment Interaction (ATI), Locus of Control, and self-efficacy suggest that a student's individual motivation, abilities, and self-efficacy are significant design considerations of instructional strategy. Specifically, the learning process would be optimized by dynamically evaluating the student's individual learning state during the training session, then adjusting the instructional intervention to increase the student's confidence and decrease anxiety. A model for dynamically tailoring instructional intervention in real-time based upon his/her individual learning characteristics and affective responses is proposed. This model describes the use of an affective component for such factors as anxiety, to be monitored and adjusted throughout the training session. The affective component interfaces with the instructor model to optimize the student-instructor interaction process (i.e. frequency of feedback, directive/reflective feedback, tone of voice). Data collection and evaluation is planned for the Conning Officer Virtual Environment (COVE), a prototypical shiphandling VE training simulator located at the Naval Air Warfare Center Training Systems Division. COVE's Intelligent Tutoring System (ITS) would benefit from the capability to provide real-time, tailored instructional intervention to the student for a variety of shiphandling tasks, students ranging from initial training for novice Ensigns to skill refreshment and mission rehearsal for expert shiphandlers, such as Commanding Officers (COs), Department Head Officers, and Division Officers. In addition, the model will be tested for the interaction of the CO with the junior officer for possible use as an affect feedback generator to the CO.