home - login - register

PERC 2008 Abstract Detail Page

Previous Page  |  New Search  |  Browse All

Abstract Title: The Difficulties in Turning Students into Numbers
Abstract: While the qualitative classification of student responses is often practiced within PER, quantitative methods open up a large set of analysis tools that are unavailable to qualitative methods.  However, since PER data is based on student responses, there are often several difficulties that arise in taking those responses and turning them into quantitative measures.  Using methods and tools from the field of classification (data mining), we look at how the type of question being asked (e.g., multiple choice, free response, ranking task) determines what kind of measurement can appropriately be made and how that in turn effects the mathematics of comparison.  We also examine the extent to which, when using a hierarchical clustering method, the resolution of quantitative ties – responses that are determined to be "equidistant" to more than one classification category – leads to changes in the final classification.
Abstract Type: Contributed Poster

Author/Organizer Information

Primary Contact: R. Padraic Springuel
University of Maine
5709 Bennett Hall
University of Maine
Orono, ME 04469
Co-Author(s)
and Co-Presenter(s)
John R. Thompson, University of Maine; Michael C. Wittmann, University of Maine