|
Location:
|
SS Ballroom ABC |
|
Date:
|
Wednesday, Aug.3 |
|
Time:
|
1:10 PM -1:20 PM
|
|
Author:
|
Steven Wolf, Michigan State University
586-484-3661, wolfste4@msu.edu
|
|
Co-Author(s):
|
Gerd Kortemeyer
|
|
Abstract:
|
On its 30th anniversary, we are re-examining the seminal paper by Chi et al., which firmly established the notion that novices categorize physics problems by "surface features" (e.g. "incline," "pendulum," "projectile motion," ...), while experts use "deep structure" (e.g., "energy conservation," "Newton 2," ...). The paper has been cited more than 3000 times in scholarly articles over a wide range of disciplines. Yet, some details of the original research design of this card-sorting experiment and its analysis methods are not clear. In replicating the study, particularly the choice of problems seems to strongly influence the outcome; only a carefully "rigged" problem set will have a good signal-to-noise ratio. We replicated the experiment with an expert group, using a large set of problems, and noted the degree to which different specific subsets of problems lead to more or less clear-cut results.
|
|
Footnotes:
|
None
|
|
|
|