Comparing Robust Variance Estimation Versus Generalized Least Squares for Spearman-Brown-Corrected Reliability Generalization of Multidimensional, Different-Length Forms

• In Event: 25.026 - Challenges in Meta-Analysis

Abstract

We compared use of GLS and RVE for multivariate synthesis of coefficient alpha estimates across studies using random versus non-random different length forms while employing the Spearman-Brown correction. The findings of this research can inform both methodological and applied meta-analysts about how GLS and RVE work for multivariate reliability generalization studies and whether Spearman-Brown’s adjustment can be used for the shortened forms.

Authors

• Luping Niu, The University of Texas at Austin

• Tasha Beretvas, The University of Texas at Austin