ABSTRACT
Questions for MILT concern the cost-effectiveness of a fully authorable tutor. It is not obvious that the minute manipulations enabled by the authoring interface are worth the computational cost. Nor is it obvious that instructors or researchers will know how to use this capability, particularly when they are faced with building sequencing rules from the ground up rather than being provided with a few top-level manipulations, such as the variable interaction modes offered by LingWorlds. The approach to authoring in MILT borrows from human factors methods of simulation modeling used in military logistics and resource planning. ICALL, by contrast, may profit from a theoretical predisposition toward the definition of tutoring variables.
