Mplus Paper Examples
Do self-report instruments allow meaningful comparisons
across diverse population groups? Testing measurement invariance using the
confirmatory factor analysis framework by Steven E. Gregorich
This paper is published in the journal
Medical Care. We are very grateful to Dr. Gregorich for writing this paper and for giving us permission to illustrate
how to obtain the results in Mplus on our web site.
The data for this paper is stored as
means and covariance matrices in the file
gregorichbw.dat. The data are from the NHANES 82-84 epidemiological
The paper shows the fit of eight
models in Table 2 and how to obtain these models using Lisrel in the Appendix.
These models are shown in Mplus below.
Model 1 -- all parameters free
- Model 2 -- metric (pattern) invariance -- loadings invariant
- Model 3 -- strong (scalar) invariance -- loadings & intercepts invariant
- Model 4 -- partial strong invariance -- tau7 free
- Model 5 -- partial strict invariance -- tau7 & theta7 free
note: strict invariance -- loadings, intercepts & residuals invariant
- Model 6 -- partial strict invariance -- tau7, theta7 * theta2 free
- Model 7 -- partial strict invariance plus equal factor means
- Model 8 -- partial strict invariance plus equal factor means & variances
The content of this web site should not be construed as an endorsement
of any particular web site, book, or software product by the
University of California.