Abstract
We propose a novel theoretical framework for knowledge transfer that consists of constructing a representation of context, generating a solution, and evaluating whether the solution makes sense given that context. Sense-making and satisficing processes determine when the transfer cycle begins and ends, and the classical mechanisms of transfer (identical rules, analogy, knowledge compilation, and constraint violation) are triggered under different representations of context. We use this framework to interpret student transfer behaviors with examples from the literature in the domains of physics and mathematics. We view this framework as a research tool to explore the dynamic nature and complexity of transfer by highlighting critical features and issues that a complete model of transfer must address. We describe how the framework relates to the classical and recent alternative approaches to understanding transfer and discuss the implications for theory development and instruction.
ACKNOWLEDGMENTS
This work was supported by Grant SBE0354420 from the National Science Foundation, Pittsburgh Science of Learning Center (http://www.learnlab.org) and Grant R305B070085 from the Department of Education, Institute of Education Sciences. No endorsement should be inferred. We thank Vincent Aleven, David T. Brookes, Clark Chinn, Stellan Ohlsson, James Greeno, Sarah Nokes-Malach, members of Cognitive Science Learning Laboratory, and three anonymous reviewers for their many helpful comments and suggestions on the paper.
Notes
In the knowledge compilation section, we describe the complementary computational mechanism in Anderson's ACT-R cognitive architecture that can account for declarative-to-procedural transfer (Anderson & Lebiere, Citation1998).
This final evaluation to determine if sense-making has been achieved can be a careful, conscientious effort encompassing the entire process or a tacit evaluation to determine if things feel OK and there are no obvious errors or contradictions.