Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.

Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Partial invariance in loadings and intercepts—their interplays and implications for latent mean comparisons

Yongwei Yang, University of Nebraska - Lincoln

Abstract

Mean and covariance structure modeling procedures have been widely used for assessing factorial invariance and the practice of partial factorial invariance (PFI) is common. This study investigates the extent to which PFI affects subsequent tests of intercept invariance and latent mean differences. The study implements a Monte Carlo experiment where factors of model size, severity of PFI, construct reliability, and intercept and latent mean difference pattern are manipulated. Results show that when partial loading invariance exists but is correctly specified, it did not negatively affect the power of the omnibus test of intercept invariance. However, results also suggest that relying on modification indices to determine noninvariant intercepts would be problematic—these indices are likely to lead to incorrect identifications of noninvariant parameters. With respect to latent mean comparisons, various forms of partial factorial invariance, when correctly specified, did not impact the test of latent mean differences. With the proliferation of educational research involving cross-group comparisons, researchers need to understand how PFI may affect analyses upon which substantive inferences are based. Findings from this study will provide a starting point for answering this very important question.

Subject Area

Educational evaluation|Statistics

Recommended Citation

Yang, Yongwei, "Partial invariance in loadings and intercepts—their interplays and implications for latent mean comparisons" (2008). ETD collection for University of Nebraska-Lincoln. AAI3297716.
https://digitalcommons.unl.edu/dissertations/AAI3297716

Share

COinS