My research concentrates on modeling of cognitive processes present in human
learning, using artificial neural networks. Studies primarily focus on the
process by which adults acquire intellectual expertise. Intellectual
expertise is known to develop in a complex and non-algorithmic manner, often
producing the elusive concept of intuitive understanding. Artificial neural
networks hold out promise for providing greater insight into this important
aspect of human learning. In the long run, many societal benefits can be
envisioned as branching off from this line of investigation, not the least
of which is the potential to improve student success in formal education.
At the Doctoral Consortium for SIGCSE2001 I presented my earliest ideas on cognitive
modeling with neural networks. I included results from my first effort at creating a
computational model of the process of learning to select solution strategies for Calculus
Integration. I also presented results
from a qualitative study that I conducted in support of this model. In the year since
that presentation, this research has evolved significantly, gone through many iterations,
and at the time of this writing I am completing the final work on this initial study.
It is my intent to share aspects of this study with Doctoral Consortium attendees. I plan to focus on data modeling issues that I have encountered, and how I have resolved them. I would not have predicted a year ago, the extent to which decisions about the selection and coding of cognitive data have become central to the success or failure of my research. As this study will be in part incorporated into my dissertation, I think it would be valuable to talk about and share the process I have gone through in learning about cognitive modeling in a connectionist environment.
Current Stage in My Studies: I will complete my third year of Doctoral Studies in December 2001. I passed my Qualifying Exams last year, shortly after SIGCSE2001. Currently I am completing this study as part of my preparation of a formal Proposal, which is tentatively scheduled for late Spring or early Fall 2002.
What I Hope to Gain from the DC: In the past, the DC has been a great venue for sharing ideas, gaining feedback from voices other than those I encounter on my own campus, and generally getting much appreciated re-energizing. This is what I hope to have happen again.
Here are a couple interesting references related to various aspects of my studies:
About Cognition and Learning with a Technology Perspective:
Duffy, Thomas M., & Jonassen, David H. "Constructivism and the Technology of Instruction" (1992), Hillsdale, NJ: Lawrence Erlbaum Associates.
An Example of Early Work Using a Neural Network to Model Higher Order Thinking:
Anderson, James A., & Spoehr, Kathryn, & Bennett, David J., "A Study in Numerical Perversity: Teaching Arithmetic to a Neural Network", in Neural Networks for Knowledge Representation and Inference. Edited by Levine, Daniel S. & Aparicio, Manuel. Hillsdale, NJ: Lawrence Erlbaum Associates. (the specifics of this reference may have changed since I originally obtained it)
About Learning College Level Mathematics in General:
Kaput, James J. & Dubinsky, Ed. "Research Issues in Undergraduate Mathematics Learning" The Mathematical Association of America, Notes Number 33 (1994)