A Monte Carlo Study of Confidence Interval Methods for Generalizability Coefficient

• In Event: AERA Virtual Poster Session 8
  In Poster Session: Poster Session 1

Abstract

Computing confidence intervals around g-coefficients has long been a challenging task in G-theory. As G-theory can be framed to a linear mixed-effect model (LMM), bootstrap and simulation techniques from LMM paradigm can be used to construct the confidence intervals. The purpose of this research is to examine four different LMM-based methods for computing the confidence intervals that have been proposed and to determine their accuracy under six simulated conditions based on the type of test scores (normal, dichotomous, and polytomous data) and data measurement design (p × i × r and p x [ i : r ]). A bootstrap technique called “parametric methods with spherical random effects” consistently produced more accurate confidence intervals than the three other LMM-based methods.

Authors

• Zhehan Jiang, Peking University

• Mark R. Raymond, none

• Ren Liu, University of California - Merced

• Dexin Shi, University of South Carolina

• Junhua Sun, Nanjing University

• Lu Liu, Peking University