KNOWLEDGE ELICITATION STRATEGIES FOR MODELING INSTRUCTIONAL EXPERTISE

THE INTRODUCTION OF ARTIFICIAL INTELLIGENCE INTO COMPUTER-BASED INSTRUCTION (CBI) PROMISED THE POSSIBILITY OF SOPHISTICATED SYSTEMS THAT COULD PROVIDE ONE-ON-ONE INSTRUCTION EFFICIENTLY AND ECONOMICALLY. UNFORTUNATELY, EFFORTS TO DEVELOP SYSTEMS THAT REPRESENT INSTRUCTIONAL EXPERTISE HAVE BEEN LARGELY DISAPPOINTING. WHETHER DUE TO THE LACK OF WELL-DEVELOPED MODELS OF INSTRUCTIONAL EXPERTISE, OR THE DIFFICULTY OF EXTRACTING KEY COMPONENTS OF THAT EXPERTISE FROM SKILLED HUMAN INSTRUCTOR, PRESENT SYSTEMS DO NOT OFFER THE SUBTLE DIAGNOSTIC SKILLS NOR FLEXIBLE RESPONSE STYLES OF EXPERT HUMAN INSTRUCTORS. KNOWLEDGE ENGINEERING METHODS ARE NEEDED THAT CAN IDENTIFY AND DOCUMENT KEY COMPONENTS OF INSTRUCTIONAL EXPERTISE, AT A LEVEL OF SPECIFICITY AND DETAIL THAT WILL SUPPORT DEVELOPMENT OF CBI SYSTEMS THAT ENCOMPASS THAT EXPERTISE. THE PROPOSED STUDY WILL EXAMINE THE FEASIBILITY OF USING CRITICAL DECISION AND CONCEPT MAPPING METHODS OF KNOWLEDGE ELICITATION TO EXTRACT AND DESCRIBE KEY FEATURES OF EXPERT ONE-ON-ONE INSTRUCTION. KNOWLEDGE ELICITATION SESSIONS WILL BE CONDUCTED WITH EXPERT INSTRUCTORS IN TWO DOMAINS (CRITICAL CARE NURSING AND COMPUTER SOFTWARE DEBUGGING). RESULTING DATA WILL FORM THE BASIS OF A TAXONOMY OF INSTRUCTIONAL EXPERTISE. THE STUDY WILL PROVIDE IMPORTANT INFORMATION ON THE FEASIBILITY OF REPRESENTING CRITICAL FEATURES OF SKILLED ONE-ON-ONE INSTRUCTION IN CBI SYSTEMS.