Please use this identifier to cite or link to this item: http://hdl.handle.net/10125/9736

Dynamic scaling : an ipsative procedure using techniques from computer adaptive testing

File Description SizeFormat 
uhm_phd_9604135_uh.pdfVersion for UH users3.89 MBAdobe PDFView/Open
uhm_phd_9604135_r.pdfVersion for non-UH users. Copying/Printing is not permitted3.94 MBAdobe PDFView/Open

Item Summary

Title: Dynamic scaling : an ipsative procedure using techniques from computer adaptive testing
Authors: Berg, Stephen R.
Keywords: Scaling (Social sciences)
Issue Date: 1995
Abstract: The purpose of this study was to create a prototype method for scaling items using computer adaptive testing techniques and to demonstrate the method with a working model program. The method can be used to scale items, rank individuals with respect to the scaled items, and to re-scale the items with respect to the individuals' responses. When using this prototype method, the items to be scaled are part of a database that contains not only the items, but measures of how individuals respond to each item. After completion of all presented items, the individual is assigned an overall scale value which is then compared with each item responded to, and an individual "error" term is stored with each item. After several individuals have responded to the items, the item error terms are used to revise the placement of the scaled items. This revision feature allows the natural adaptation of one general list to reflect subgroup differences, for example, differences among geographic areas, ethnic groups, and socioeconomic groups. It also provides easy revision and limited authoring of the scale items by the computer program administrator. This study addressed the methodology, the instrumentation needed to handle the scale-item administration, data recording, item error analysis, and scale-item database editing required by the method, and the behavior of a prototype vocabulary test in use. Analyses were made of item ordering, response profiles, item stability, reliability and validity. Although slow, the movement of unordered words used as items in the prototype program was accurate as determined by comparison with an expert word ranking. Person scores obtained by multiple administrations of the prototype test were reliable and correlated at .94 with a commercial paper-and-pencil vocabulary test, while holding a three-to-one speed advantage in administration. Although based upon self-report data, dynamic scaling instruments like the model vocabulary test could be very useful for self-assessment, for pretests and readiness measures, and for low-stakes placement tests. Dynamic scaling appears to be a potentially useful tool for the development and administration of quick and accurate scale instruments.
Description: Thesis (Ph. D.)--University of Hawaii at Manoa, 1995.
Includes bibliographical references (leaves 162-166).
Microfiche.
166 leaves, bound ill. 29 cm
URI/DOI: http://hdl.handle.net/10125/9736
Rights: All UHM dissertations and theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission from the copyright owner.
Appears in Collections:Ph.D. - Educational Psychology



Items in ScholarSpace are protected by copyright, with all rights reserved, unless otherwise indicated.