Progressions of Qualitative Models as a Foundation for Intelligent Learning Environments
Abstract
The design of our intelligent learning environmental is based upon a theory of expertise and its acquisition. We find that when reasoning about physical systems, experts utilize a set of mental models. For instance, they may use qualitative as well as quantitative models, and behavioral as well as functional models. The transition from novice to expert status can be regarded as a process of model evolution: students formulate a series of upwardly compatible models, each of which is adequate for solving some subset of problems within the domain. Further, students need to evolve not just a single model, but rather a set of models that embody alternative conceptualizations of the domain. Finally, we claim that in the initial stages of learning, students should focus on the acquisition of qualitative models: quantitative models should be introduced only after the domain is understood in qualitative terms. This article focus primarily on qualitative, behavioral models of electrical circuit operation designed to make the casuality of circuit behavior derive clearly from basic physical principles. The constraints on model evolution, in terms of causal consistency and learnability, are discussed and a sequence of models that embody a possible transformation from novice to expert status is outlined.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 01, 1986
- Accession Number
- ADA169725
Entities
People
- Barbara Y. White
- John R. Frederiksen
Organizations
- BBN Technologies