Labeling variance of an oberved variable PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Xu, Man posted on Friday, October 24, 2008 - 11:15 am
Hello, I am trying to cacluate a effect size, but apparently I couldn't get it run. I think it's the labeling that I used caused the problem. Could you help out please? Thank you very much!

MODEL:
ZST37Q01 ON ZPVEPS(b1)
ZPVISI(b2)
ZPVUSE(b3);

ZPVEPS(var1);
ZPVISI(var2);
ZPVUSE(var3);
ZST37Q01(Theta);

MODEL CONSTRAINT:

new(ES1);
ES = b1*(2*sqrt(var1)/sqrt(var1*b1**2+Theta));
 Xu, Man posted on Friday, October 24, 2008 - 11:29 am
Sorry, it's a labeling and typing mistake indeed. I have corrected it to be

ES1 = b1*(sqrt(var1)/sqrt(var1*b1**2+Theta)

But the estimate for ES1 is 0.476, whereas the STDYX standardization output is 0.529, and their t values are 17.190 and 13.814.

Could you advice me why is this please? Thanks!
 Linda K. Muthen posted on Friday, October 24, 2008 - 7:08 pm
I think your problem is that theta is a residual variance not a variance.
 Xu, Man posted on Friday, October 24, 2008 - 7:17 pm
Oh i see. thank you. i forgot to add explained variances of the other two variables. i will take another try.
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: