The Development of an Automated Cognitive Task Analysis and Modeling Process for Intelligent Tutoring System Development
Abstract
Studies have shown that 98% of students with private tutors do better than students in the traditional classroom. This has initiated research and the development of computerized intelligent tutoring systems based upon research related to cognitive learning and the communication of knowledge between a tutor and a student. Intelligent tutoring systems have demonstrated a 65% improvement in what can be learned in a given period of time as compared to traditional computer aided instruction in Navy tasks. The principle component of these computerized tutors is a model of the knowledge which makes up the cognitive task to be learned by a student. Creating these cognitive models requires an extensive analysis of the task knowledge. This process is expensive yet critical to the success of these tutors. This research has produced a software tool which interacts with a subject matter expert and conducts an analysis of his/her knowledge, building the needed cognitive task model for use in intelligent tutoring system development. An evaluation of the system was conducted to determine the degree of variation in models generated by experts on a specific highly proceduralized task. The subject generated models were also compared to a model generated by a cognitive analyst. The results demonstrated a high degree of consistency in the models developed between subjects although subjects tended to organize their knowledge of procedures into units of differing size. Cognitive task analysis, Cognitive simulation modeling intelligent tutoring systems, GOMS, Human systems interaction transfer of training, Training time prediction.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1993
- Accession Number
- ADA269782
Entities
People
- Kent E. Williams
Organizations
- Virginia Tech