Accounting for Errors in Model Analysis Theory: A Numerical Approach Documents
Steven Sommer and
Rebecca S. Lindell
By studying the patterns of a group of individuals' responses to a series of multiple-choice questions, researchers can utilize Model Analysis Theory to create a probability distribution of mental models for a student population. The eigenanalysis of this distribution yields information about what mental models the students possess, as well as how consistently they utilize said mental models. Although the theory considers the probabilistic distribution to be fundamental, there exists opportunities for random errors to occur. In this paper we will discuss a numerical approach for mathematically accounting for these random errors. As an example of this methodology, analysis of data obtained from the Lunar Phases Concept Inventory will be presented. Limitations and applicability of this numerical approach will be discussed.
- Download PERC03_Sommer.pdf - 144kb Adobe PDF Document
Published September 9, 2004
Last Modified July 7, 2013
This file is included in the full-text index.