MODEL-BASED PROTOCOL FOR OPTIMIZED LEARNING

THE PRIMARY OBJECTIVE OF THIS STUDY IS TO EVALUATE THE FEASIBILITY OF DEVELOPING A MODEL-BASED PROTOCOL FOR OPTIMIZED INSTRUCTION. THE BASIC TECHNIQUE CENTERS ON THE USES OF THE OPTIMAL CONTROL MODEL (OCM) OF THE HUMAN OPERATOR, AN INFORMATION-PROCESSING MODEL OF SKILLED HUMAN BEHAVIOR WHOSE STRUCTURE AND PARAMETER SET REFLECT THE BASIC HUMAN LIMITATIONS AND COMPONENT SKILLS WHICH DETERMINE TASK PERFORMANCE. WE PROPOSE USING A PARAMETRIC LEARNING MODEL TO SPECIFY THE TIME HISTORY OF MODEL PARAMETERS DURING TRAINING, FOR BOTH INDIVIDUAL SUBJECTS AND NORMAL POPULATIONS. LEARNING DEFICIENCIES ARE DETECTED AND DIAGNOSED BY COMPARING INDIVIDUAL VS. POPULATION PARAMETER TIME HISTORIES. SKILL-DEFICIENT INDIVIDUALS UNDERGO REMEDIAL TRAINING ON TAKS SELECTIVELY TUNED TO AMELIORATE THE SPECIFIC DEFICIT; SUCCESSFUL REMEDIAL TRAINING RETURNS THE INDIVIDUAL TO THE MAIN INSTRUCTIONAL SEQUENCE. WE PROPOSE TO EVALUATE OVERALL PROTOCOL FEASIBILITY VIA FOUR TASKS. WE WILL: 1) EXTEND THE PARAMETRIC LEARNING MODEL TO SUPPORT STATISTICAL DECISION-MAKING; 2) SPECIFY A METHOD FOR DEFICIT DIAGNOSIS AND DESIGN SELECTIVELY-SENSITIVE REMEDIAL TASKS; 3) EVALUATE THE PROTOCOL WITH A BASELINE TASK AND EXPERIMENTAL PROGRAM; AND 4) IDENTIFY PROTOCOL DEFICIENCIES AND SPECIFY REQUIREMENTS FOR FOLLOW-ON DEVELOPMENT. A FINAL REPORT WILL SUMMARIZE THE STUDY'S OBJECTIVES, ACHIEVEMENTS, AND RECOMMENDATIONS FOR FURTHER WORK.