Evaluation of the Computer Aided Training Evaluation and Scheduling (CATES) Decision Model for Assessing Flight Task Proficiency.

The efficacy of the CATES system for making training decisions and determining student proficiency in Naval in-flight training proposed in an earlier study (Rankin and McDaniel, 1980) is compared with the present system of instructor judgments for performance assessment. The current study used 29 newly-designated naval aviators undergoing Fleet Replacement Pilot Training in the SH-3 aircraft. From an inventory of 190 tasks, 18 tasks were selected to evaluate the model. Standard training materials and equipment were used, and performance was graded as students proceeded through the training syllabus. The task performance information required to reach a decision and the level of student proficiency upon completion of the training program were then analyzed. Results indicate the CATES system requires less information to make a decision than the current human-judgment system and is reliably more accurate, suggesting greater consistency and accuracy of mathematical models to an actual training situation in a considerably more unstructured environment than previous research studies. This report also provides extensive details of the statistical decision model used by the CATES system, results of other evaluations using mathematical models versus human models in decision making, study definitions, research methods, results, and conclusions. Six appendices include related materials and 30 references. (LMM)