Covariance PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Marike Deutz posted on Friday, June 30, 2017 - 9:18 am
I am estimating a bifactor model, and want to regress two independent variables (iv1 iv2) on the factors in the model. Furthermore, I want to control for covariates (cv1 cv2). Simplified version of my syntax:

f1 BY i1 i2 i3 i4 i5;
f2 BY i6 i7 i8 i9 i10;
f3 BY i11 i12 i13 i14 i15;
bifactor BY i1-i15;

f1 f2 f3 bifactor ON cv1 cv2 iv1 iv2;

I also want to control for relations between my ivs and covariates and among my ivs. However I do not want to control for cv3 in the regression model and also not for the covariance between cv1 and cv2. I also need WITH statements to include my full sample.

Should I add:
iv1 iv2 ON cv1 cv2 cv3;
iv1 WITH iv2;

Or:
iv1 WITH cv1 cv2 cv3;
iv2 WITH cv1 cv2 cv3;
iv1 WITH iv2;
cv1 WITH cv2@0;


Also: would you recommend standardizing ivs and cvs? Also binary variables? Some of my models only work with standardized variables.

Thank you so much for your help.
 Marike Deutz posted on Friday, June 30, 2017 - 9:49 am
I know posts should be limited to 1 but I just realized it's important information that I used the WLSMV estimator because of categorical indicators (and my cv2 and cv3 are binary).
 Bengt O. Muthen posted on Friday, June 30, 2017 - 5:04 pm
I would not use either of your 2 additions.

You may also want to ask on SEMNET.
 Marike Deutz posted on Monday, July 03, 2017 - 6:46 am
Thank you. I will ask somewhere else.
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: