AN AUTOMATED TRAINING SYSTEM FOR BUILDING EXPERTISE

COMPUTER-BASED TUTORIAL SYSTEMS HAVE NOW BECOME EFFECTIVE TRAINERS IN A NUMBER OF DOMAINS. MANY GOOD SYSTEMS ARE MODELED AFTER THEORIES OF HUMAN LEARNING AND KNOWLEDGE REPRESENTATION. HOWEVER, SUCH SYSTEMSTYPICALLY ASSUME A LIMITED VIEW OF HUMAN KNOWLEDGE, HENCE LIMITING THEIR APPLICABILITY TO INTRODUCTORY LEVEL MATERIAL. RESEARCH BY THE PROPOSED PROJECT TEAM HAS SHOWN THAT EXPERTS HAVE A BROAD RANGE OF KNOWLEDGE STRUCTURES. ANY TRAINING SYSTEM (COMPUTER OR OTHER) ATTEMPTING TO BUILD STUDENT'S EXPERTISE SHOULD GUIDE ITS INSTRUCTION ACCORDING TO HOW EXPERTS REPRESENT AND USE KNOWLEDGE. WE PROPOSE TO BUILD A COMPUTER-BASED TRAINING SYSTEM AND OVERALL TRAINING METHODOLOGY THAT HELPS STUDENTS INTEGRATE A WIDE RANGE OF KNOWLEDGE (E.G., PLANNING, CAUSAL REASONING, CONCEPT FAMILIARITY) AND APPLY IT TO PRACTICAL PROBLEMS. IN ADDITION, WE PROPOSE TO USE KNOWLEDGE ELICITATION TECHNIQUES DEVELOPED BY THE PROPOSED PROJECT TEAM TO EVALUATE THE EFFECTIVENESS OF TRAINING AND GUIDE CORRECTIVE INSTRUCTION.