Which Model is Best? Robustness Properties to Justify Model Choice Among Unidimensional IRT Models under Item Parameter Drift

Authors

  • André A. Rupp
  • Bruno D. Zumbo

DOI:

https://doi.org/10.11575/ajer.v49i3.54984

Abstract

This article extends recent research on item parameter drift by investigating the robustness properties of basic unidimensional IRT models. Specifically, the article explores whether it is possible to advocate the consistent choice of one model over another based on its robustness properties under drift. On the one hand, it is shown that the biases that are introduced due to drift are minor for most practically relevant circumstances across all models. On the other hand, it is shown that the mathematical structure of the biases is theoretically complex so that globally superior performance of one model over another is observed only under restrictive side conditions.

Downloads

Published

2003-10-01

Issue

Section

ARTICLES