Handling Covariates PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Luci M. posted on Wednesday, May 11, 2011 - 6:03 pm
Hello!

With the resources you gave me, I was able to figure out the answers to my previous questions.

I now have a technical question -- when running my SEM model, I am getting the estimates. However, when I use the WITH option to corelate my covariate, I am no longer getting the SE and p values on my output. I am only getting the Estimates. Even though the Estimates don't seem to change, I am wondering why this is.

Should I even use the WITH command in the first place or is MPlus correlating the covariates by default? I am not sure if the correlation between the covariates is taken into account when MPlus produces the Estimates.

Thanks very much.
 Linda K. Muthen posted on Thursday, May 12, 2011 - 9:21 am
If you do not get standard errors, the model did not converge.

You should not include WITH statement for the covariates. The model is estimated conditioned on these variables. If you want there values, do a TYPE=BASIC.
 Katy Roche posted on Thursday, September 17, 2020 - 7:37 am
I brought in the means for the Xs to deal with missing. However, when I do this, my Xs (manifest covariates) are correlated. When I set those correlations to 0, fit drops considerably. Any suggestions?
 Bengt O. Muthen posted on Thursday, September 17, 2020 - 6:07 pm
Your Xs should be correlated. They are implicitly correlated in a regular regression model.
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: