Abstract

The interpretive processes required to understand the context and goals of an algorithmic task are illustrated in the use of an intelligent instructional system developed to train soldiers to monitor a computerized missile's system automatic identification of aircraft. The problems students had in understanding the identification task were addressed in INCOFT, a simulation-based intelligent instructional system that depends, in part, on human instructors to convey the task framework. Supported by recent advances in the cognitive science of instruction, the concept ofapprenticeship provides a model for the student-machine as well as the student-instructor interactions observed in the use of INCOFT. The system's instantiation of apprenticeship illustrates an alternative to conventional tutorial-based design of intelligent instructional systems