Version 8.10 of Mplus is now available on the OIT Windows Terminal Server Apps@UT. Mplus analyzes Structural Equations Models (SEM) that can include both continuous and categorical latent constructs. According to the Mplus website, new features of this version of Mplus include: 

ALIGNMENT  

The ALIGNMENT option is used with multiple group and longitudinal models to assess measurement invariance and compare factor means, variances, and structural parameters across groups and time. It is available for multiple group and longitudinal models when all variables are continuous or binary with the ML, MLR, MLF, and BAYES estimators and when all variables are ordered categorical (ordinal) with the ML, MLR, MLF, and WLSMV estimators. 

TESTING FOR MEASUREMENT INVARIANCE 

The MODEL option of the ANALYSIS command is used to automatically test multiple group models and single-group longitudinal models for measurement invariance. For multiple group models, the GROUPING option or the KNOWNCLASS option is used. Measurement invariance testing is available for CFA and ESEM models for continuous variables using the maximum likelihood and Bayes estimators; for censored variables using the weighted least squares and maximum likelihood estimators; for binary and ordered categorical (ordinal) variables using the weighted least squares, maximum likelihood, and Bayes estimators; and for count variables using the maximum likelihood estimator. It is not available for censored-inflated, count-inflated, nominal, continuous-time survival, negative binomial variables, or combinations of variable types. 

PENALIZED STRUCTURAL EQUATION MODELING  

A new method called Penalized Structural Equation Modeling (PSEM) is available. Penalized Structural Equation Modeling (PSEM) uses a prior to impose a penalty on model estimation. The following four penalty functions are available for PSEM: Normal (N), Alignment Loss Function (ALF), LASSO, and GEOMIN. They can be used with the ML, MLR, and WLSMV estimators to improve estimation of models like EFA, SEM, growth models, multiple group alignment, and longitudinal alignment. PSEM for categorical variables uses weighted least-squares estimation.  

RANDOM CORRELATIONS  

Random correlations are available for continuous, binary, ordered categorical (ordinal), or combinations of these variable types for TYPE=TWOLEVEL and TYPE =CROSSCLASSIFIED. The | symbol is used in conjunction with TYPE=RANDOM and ESTIMATOR=BAYES to name and define the random correlation variables in the model. The name on the left-hand side of the | symbol names the random correlation variable. The variables on the right-hand side of the | symbol specify the variables that will have a random correlation. 

Availability and Assistance

The OIT Windows Terminal Server Apps@UT (apps.utk.edu) is available at no additional charge for students, faculty, and staff on the UT Knoxville campus. OIT offers Minimal Support for Mplus to University of Tennessee, Knoxville-area students, faculty, and staff for this software.

For assistance, contact the OIT HelpDesk at 865-974-9900.