Investigating Separate and Concurrent Approaches for Item Parameter Drift in 3PL Item Response Theory Equating. Issue 1 (2nd January 2017)

• Record Type: Journal Article
• Authors: Arce-Ferrer, Alvaro J.; Bulut, Okan
• Is Part Of: International journal of testing. Volume 17:Issue 1(2017)
• Page Start: 1
• Page End: 22
• Publication Date: 2017-01-02
• Subjects: 3PL equating -- item parameter drift -- item response theory -- scaling; Psychological tests -- Periodicals; Educational tests and measurements -- Periodicals; 150.28705
• DOI: 10.1080/15305058.2016.1227825
• Languages: English
• ISSNs: 1530-5058

Abstract

Abstract : This study examines separate and concurrent approaches to combine the detection of item parameter drift (IPD) and the estimation of scale transformation coefficients in the context of the common item nonequivalent groups design with the three-parameter item response theory equating. The study uses real and synthetic data sets to compare the two approaches based on IPD flagging rates, type I error and power rates, and recovery of scale transformation coefficients. Results indicate that the two approaches render similar outcomes with stable anchor sets. However, they can produce dissimilar results with unstable anchor sets because of differences in the performance of their IPD components. Further, the findings of this study caution about working backward from equated cut scores to motivate the selection of an anchor set.