Bayesian SEM; MODEL CONSTRAINT: PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Dex  posted on Thursday, March 10, 2016 - 6:40 pm
Hello, I have a question about the "MODEL CONSTRAINT" option under "estimator=bayes;" For example, consider a SEM model specified as

f BY x1(A1)
x2*(B1)
x3*(C1)
f@1;
[x1*](D1);
[x2*](E1);
[x3*](F1);
x1*(G1);
x2*(H1);
x3*(I1);


If I use
MODEL CONSTRAINT:
new(z1);
Z1=0.2*A1+D1
I would assume that the posterior of Z1 could be interpreted as the posterior of the expected value for X1 when Theta=0.2. (X1_hat=theta*lamda1+tau1).

However, I was wondering could the residual term be incorporated using the same manner. Given X1=theta*lamda1+tau1+epsilon1. If I specify
MODEL CONSTRAINT:
new(z2);
Z2=0.2*A1+D1+G1, would z2 can be interpreted as the posterior for X1 when theta=0.2?
More specifically,
In Z2=0.2*A1+D1+G1, will "G1" be treated as the residual "variance"(which is actually estimated in the model) or the actual residual term by Mplus?
 Tihomir Asparouhov posted on Friday, March 11, 2016 - 8:56 am
It is the residual variance.
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: