Overall, the three demonstrations of automated adaptive training ranging from basic instrument flight training tasks to air-to-air intercept tasks were successful. The data indicates that:

• Training economies can be realized,

• Student motivation can be increased,

• Training scope can be expanded,

• Objective performance data can be developed, and

• Unburdening of the instructor can be achieved.

The latter point is considered particularly important since the instructor can be freed to observe the student's behavior and provide the necessary assistance. He need no longer devote full time to operation of a complex trainer, much less managing and directing the syllabus through a maize of switches, knobs and dial settings. Of more importance, however, was the general result that existing training knowledge can be brought to bear on simulation training problems with potentially significant impact. There is little doubt but that adaptive techniques based on objective performance measurement can decrease training time and improve learning. The instructors can be unburdened and permitted to help shape the student's behavior. Uniform training output can be achieved by requiring objective demonstration of criterion performance.