Case-Based Reasoning for Engineering Statistics

Arizona State University, Scottsdale AZ

Investigators

Abstract

Using findings from cognitive science, a prototype of an intelligent tutoring system (ITS) is being developed. This ITS utilizes case-based reasoning (CBR) to scaffold undergraduate engineering students in their learning of introductory probability and statistics. This scaffolding process is based on the principle that successful problem solvers and experts gain a better grasp of the underlying problem structure by moving back and forth as they reach dead-ends or see new ways to formulate the problem. Specifically, the students learn the following three steps: 1) situation conception--understand the problem in everyday language; 2) mathematized situation conception--create a mathematical representation; and 3) solution method conception--carry out the mathematical procedures. Existing commercial software and well-established research index-based CBR and rule-based expert systems are used to support the presentation of ITS by graduate students to the undergraduate students in the course. The assessment of ITS is to test two hypotheses: 1) Does the ITS increase the internalization of key concepts by the students? and, 2) Does the ITS increase the student independence in representing and solving problems? Also, assessments based on comparing the performances on questions related to the hypotheses for ITS-students and non-ITS-students. This initial evaluation of ITS from the standpoint of integrating it with an existing engineering statistics course, allows the participation of other universities who also offer the engineering statistics course to participate in a joint, longer term project to realize a fully functional ITS. Also, the partnership with other universities provides a vehicle for embedding ITS into other courses where the same cognitive and instructional principles are appilcable. The final product--a prototype and recommendations for a more fully functional ITS-- will: 1. Assist students in extracting the underlying common structure from engineering statistics problems that illustrate the full range of engineering disciplines; 2. Allow the student to generate, customize, and change a virtual infinite collection of exercises that can be solved with the assistance of the ITS. The students can explore the effect of changes to different parameters and how those changes influence the solution; and 3. Help students formulate and solve practical and open-ended problems. The final product will be kept on CD-ROM, and further dissemination will be via workshops. Keywords: Case-based reasoning, intelligent tudoring system