|
Message/Author |
|
adwin posted on Thursday, January 29, 2015 - 6:42 pm
|
|
|
Good afternoon, I am currently conducting a data estimation that involve some latent variables with ordinal-scale observed variables and also ordinal (1-7 Likert) dependent variable. I use Generalised SEM of STATA 13 to estimate my model. However, I encounter a problem especially when I need to test the 'goodness of fit' and 'indirect effect', as STATA does not have such test instruments for its GSEM. I am wondering if MPLUS can solve my problem. If so, I am happily to move to MPLUS. Kind Regards, |
|
|
Indirect effects can be obtained by Mplus. With categorical variables, Mplus provides bivariate tests of fit (TECH10) and in addition, the WLSMV estimator gives an overall goodness of fit test. |
|
Back to top |
|
|