Adding chi-square test for the output PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Hadas Hawlena posted on Sunday, May 06, 2012 - 7:35 am
Dear Drs. Muthen,
I am running path analyses where the dependent are categorical and the independents could be nominal, categorical, or continuous, using:
TYPE = random;
ALGORITHM = INTEGRATION;
ESTIMATOR = MLR
OUTPUT:
SAMPSTAT
Is there a way to add to the output chi-square results of goodness of fit?
Thanks, Hadas
 Linda K. Muthen posted on Sunday, May 06, 2012 - 5:17 pm
No, chi-square and related fit statistics are not available unless means, variances, and covariances are sufficient statistics for model estimation.
 Hadas Hawlena posted on Monday, May 07, 2012 - 12:29 pm
Does it mean that I can rely only on the model selection (e.g. AIC) values to quantify the relative fit of the competing models to the data, or can I estimate the model in a different way so I can get fit statistics? If so, which model definitions should I change?
Many thanks, Hadas
 Linda K. Muthen posted on Monday, May 07, 2012 - 5:10 pm
You can use WLSMV instead of ML. You will obtain probit regression instead of logistics regresson.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: