NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED509922
Record Type: Non-Journal
Publication Date: 2010-May-3
Pages: 32
Abstractor: As Provided
Reference Count: 37
ISBN: N/A
ISSN: N/A
Do Vertical Scales Lead to Sensible Growth Interpretations? Evidence from the Field
Briggs, Derek C.
Online Submission, Paper presented at the Annual Meeting of the American Educational Research Association (Denver, CO, Apr 30-May 4, 2010)
The purpose of creating a developmental (i.e., vertical) score scale is to make it possible to compare students in terms of differences in magnitude. In other words, the scale should have interval properties. When this has been accomplished one can claim to measure growth in a sense that is both coherent and meaningful. In this paper I argue that the practice of creating and maintaining developmental scales for large-scale assessment systems in the United States is at odds with this purpose, and that this is due to confusion over the theoretical rationale under which item response theory calibration leads to scales with interval properties. The problem seems to stem from a conflict in the philosophical underpinnings of what it means (and what it requires) to support the claim that one has measured a latent construct, an issue that has been brought to light by the recent work of Joel Michell and Denny Borsboom. Using data drawn from 14 states with established developmental scales for math and reading, I illustrate the equivocal nature of aggregate growth interpretations when one follows the implicit premise that--despite claims to the contrary--developmental scales communicate only ordinal information. (Contains 1 table, 5 figures and 0 footnotes.)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Identifiers - Location: United States