NotesFAQContact Us
Search Tips
Back to results
ERIC Number: ED529215
Record Type: Non-Journal
Publication Date: 2011
Pages: 198
Abstractor: As Provided
ISBN: ISBN-978-1-1246-4982-5
The Impact of Linking Distinct Achievement Test Scores on the Interpretation of Student Growth in Achievement
Airola, Denise Tobin
ProQuest LLC, Ph.D. Dissertation, University of Arkansas
Changes to state tests impact the ability of State Education Agencies (SEAs) to monitor change in performance over time. The purpose of this study was to evaluate the Standardized Performance Growth Index (PGIz), a proposed statistical model for measuring change in student and school performance, across transitions in tests. The PGIz is a modification of the Performance Growth Index (PGI, Mulvenon, Stegman, McKenzie, Airola & Daniel, 2004). PGIz and PGI scores were calculated for a cohort of 77,812 students with four continuous years of statewide test scores. Scores from two distinct Norm-Referenced Tests (NRTs) and scores from an SEA Criterion-Referenced Test (CRT) were provided by a Midwestern SEA for 2006 to 2009. Limits of Agreement (LOA) and Kappa Coefficients were calculated at the student, school and district levels to assess the level of agreement and classification consistency for two comparisons: PGIz compared to the original PGI for the two years where the same test was employed and PGIz using the NRTs that transitioned compared to PGIz using the CRT that did not transition. The PGIz and original PGI demonstrated almost perfect agreement in classification consistency (k = 0.89-0.90) for the years where the same tests were employed. In contrast, the classification consistency of PGIz using the NRTs in transition and the PGIz using CRT scores that did not transition demonstrated very poor agreement (k = 0.07-0.25). Further, the LOA statistics revealed a bias in difference scores suggestive of anomalies in the horizontal equating of the CRT, rather than the functioning of the PGIz. The PGIz functioned as a stable indicator of performance change using the NRT exams, even across the transition, particularly for PGIz scores within three standard deviations of the mean. Recommendations include the use of the PGIz for tests in transition that employ a nationally representative external reference group for situating students initial and subsequent position in the distribution, and at the student level, extreme PGIz scores should be supported with additional evidence to avoid unintended consequences for students. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page:]
ProQuest LLC. 789 East Eisenhower Parkway, P.O. Box 1346, Ann Arbor, MI 48106. Tel: 800-521-0600; Web site:
Publication Type: Dissertations/Theses - Doctoral Dissertations
Education Level: Elementary Secondary Education
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A