ULS Estimator PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Jeffrey Hall posted on Friday, May 05, 2006 - 5:39 am
Good morning,

I have a question about the use of the ULS etimator when running a model of the following nature: type = general missing complex h1. The model that I am evaluating converges however several problems are evident:
1) There is no chi-square value for the test of model fit (value =.000, degrees of freedom =2, p-value =1.000);
2) The CFI and TLI both =1.000; RMSEA =.000;
3)Standard errors for factor loadings of the factor used as the central dependent variable are not reported but replaced with *******; accordingly the est/se contains .0000 providing no information about the statistical test.
4) The previous situation is repeated with some but not all regression coefficients (approximately 4 of 9 ) for factors upon which the dependent variable is regressed. (This issue is apparent only for the dependent variable)

Do you have any thoughts on the source of these issues? I have attempted to estimate this model using the WLSMV estimator, but have been unsuccessful as of this message. Any input on this is greatly appreciated. Thanks!
 Bengt O. Muthen posted on Friday, May 05, 2006 - 5:46 am
SEs that have huge values are shown as *****. Getting huge SE values implies a non-identified model. To see why that is so, you would have to send your input, output, data and license number to support@statmodel.com.
 Jeffrey Hall posted on Friday, May 05, 2006 - 6:40 am
Thanks for the information. I am evaluating the model for problems and will get back to you shortly.
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: