Redo Winbugs analysis in Mplus - Medi... PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Michel van der Borgh posted on Thursday, August 06, 2015 - 10:06 am
Does anyone know whether it is possible to test this Winbugs mediation model in Mplus?

I think it is a mixture model given that V[i] is a categorical latent variable.

model {
for (i in 1:N) {
mu.y[i] <- beta[1]+beta[2]*mu.m[i]+beta[3]*z[i,1]+gamma*x[i,1]+rho*(m[i]-mu.m[i])
mu.m[i] <- alpha*x[i,1]+lambda[ V[i] ]
y[i] ~ dnorm(mu.y[i],xsi[1]) # Distribution of Y
m[i] ~ dnorm(mu.m[i],xsi[2]) # Distribution of M
V[i] ~ dcat(p[]) # Latent discrete instruments
}
p[1] ~ dbeta(1,1) # Prior for the latent instrument probabilities
p[2]<-1-p[1]
lambda[1] ~ dnorm(0, 0.0001) # Distribution of the latent instrument coefficients
lambda[2] <-lambda[1] + nu
}
 Bengt O. Muthen posted on Friday, August 07, 2015 - 5:38 pm
Yes, looks like the mediator is given a 2-class mixture distribution. Using mixtures in mediation modeling was discussed under the heading Nominal mediator in the paper on our website:

Muthén, B. (2011). Applications of causally defined direct and indirect effects in mediation analysis using SEM in Mplus. Click here to download the paper. Click here to view the Technical appendix that goes with this paper and click here for the Mplus input appendix. Click here to view Mplus inputs, data, and outputs used in this paper.

Bayes can be used for mixtures in Mplus.

It would be best for you to first get the statistical formulas giving the background for the Winbugs code.
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: