An emphasis on self-development in the military community has highlighted the need for adaptive computer-based tutoring systems (CBTS) to support point-of-need training in environments where human tutors are either unavailable or impractical. Effective human tutors ask questions, tailor feedback, provide opportunities for reflection, and change the content, direction, pace, and challenge level of instruction to optimize learning (e.g., acquisition of knowledge or skills). Adaptive CBTS also attempt to select optimal instructional strategies to meet the specific learning needs of individuals or teams. To make these optimal instructional strategy decisions, the adaptive CBTS assesses trainee attributes (e.g., progress, behaviors or physiology), uses these attributes to classify states and predict learning outcomes (e.g., performance, skill acquisition, retention), and then adapts the instruction to influence learning. A truly adaptive CBTS must have a suitable trainee model, a repertoire of instructional strategies, and a methodology for selecting the best strategy. Significant challenges in the design and development of adaptive CBTS include methodologies to: assess the influence of trainee attributes that inform positive/negative learning states (e.g., confusion, boredom, frustration, and pleasure); and assess the influence of specific instructional strategies on learning given the learner's state and the training context (e.g., tasks, conditions, and learning objectives). This paper considers a modular tutoring system framework to support the authoring and assessment of adaptive tutoring capabilities. The Generalized Intelligent Framework for Tutoring (GIFT) supports authoring standards and allows users to manipulate models, libraries, and domain-specific content to empirically determine the influence of variables of interest (e.g., learning style, sensor data, feedback modes, and stress) on learning. The framework supports a variety of experimental views, including ablative tutor studies, tutor vs. traditional classroom training comparisons; evaluation of intervention vs. non-intervention strategies; pedagogical model comparisons; and tutor vs. tutor comparisons.