RMSEA in CFA using WLSMV estimation PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
Message/Author
 Anonymous posted on Tuesday, December 14, 2004 - 6:53 pm
I've been running a number of CFA models on the same data set using dichotomous indicators and the WLSMV estimator. Models range between 1-3 latent variables. I noticed that following the analysis on a couple of the models the RMSEA was not reported. On the models concerned CFI=1 and TLI>1. Be glad of an explantion for this. Many thanks indeed.
 Linda K. Muthen posted on Wednesday, December 15, 2004 - 7:41 am
I don't know which version of Mplus you are using. I think we now print something for RMSEA in all cases. Could you send the output and data to support@statmodel.com. RMSEA is most likely zero in your case.
 Tan Teck Kiang posted on Sunday, June 03, 2007 - 10:57 pm
While a simple CFA model with 4 indcators output from MPlus shows a RMSEA of 0.149 but LISREL is 0.019. Is the definition of RMSEA of MPlus different from that of LISREL. Thanks.
 Linda K. Muthen posted on Monday, June 04, 2007 - 10:45 am
Are the estimators the same? Are the chi-square values the same? Are the estimators the same?
 Brent Donnellan posted on Friday, June 25, 2010 - 9:27 am
Does the new version of Mplus provide 90% CIs for the RMSEA when models are estimated with WLSMV? An editor is requesting that we report this information for a paper. Does anyone have a good response to her/his request?

Thank you!
 Linda K. Muthen posted on Friday, June 25, 2010 - 10:02 am
Yes, Version 6 has 90% confidence intervals for WLSMV.
 Erin Kube posted on Tuesday, March 18, 2014 - 12:52 pm
Is there a separate command to produce confidence intervals for RMSEA using WLSMV. I conducted a CFA model using version 6 with WLMSV and it did not produce CIs for RMSEA.

Thanks in advance!
 Linda K. Muthen posted on Tuesday, March 18, 2014 - 1:31 pm
No, there is no option for this.
 Lauren Brumley posted on Wednesday, July 18, 2018 - 10:31 am
I also had an editor ask for confidence intervals for RMSEA using WLSMV. I am using version 6. To confirm, there is no option to request CIs for RMSEA using WLSMV, correct?

Thank you!
 Tihomir Asparouhov posted on Wednesday, July 18, 2018 - 12:59 pm
It is available in the current version.
 Lauren Brumley posted on Wednesday, July 18, 2018 - 1:21 pm
Ok. I have to use TYPE=COMPLEX because I have a stratification variable. Is that why I'm not getting the confidence intervals for RMSEA with WLSMV?

I do get CIs for RMSEA with TYPE=COMPLEX when indicators are continuous, just not categorical (WLSMV).

Thank you very much for your help.
 Tihomir Asparouhov posted on Wednesday, July 18, 2018 - 2:57 pm
You need Mplus 8.1
 Lauren Brumley posted on Wednesday, July 18, 2018 - 5:03 pm
Ah, got it, makes sense. Thank you!
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: