Asparouhov and Muthen ASA SSRM (2006)... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
 James Algina posted on Tuesday, May 29, 2018 - 9:03 am
In the second column of page 2719 of referenced paper estimation of the asymptotic covariance matrix is described for an example with stratified psus, suss and tsuss, and weights at the secondary (cluster) and tertiary levels (individual). I have two questions:

1. Is this a description of the procedure called complex two-level in Mplus, albeit with stratification and weights?

2. If there is no stratification and no weights, is the likelihood rather than a pseudo likelihood maximized.


 Tihomir Asparouhov posted on Tuesday, May 29, 2018 - 9:55 am
1. Yes, type=complex twolevel

2. Yes, if there are no weights, the weights become 1, and the pseudo log-likelihood becomes the log-likelihood.
 James Algina posted on Tuesday, May 29, 2018 - 8:05 pm
Thank you for your helpful reply.

To follow up: Does Equation (5) on page 425 of Sampling Weights in Latent Variable Modeling (Asparouhov, SEM, 2005) provide the formula for asymptotic covariance matrix for type = complex?

If not can you provide a reference?


 Tihomir Asparouhov posted on Wednesday, May 30, 2018 - 8:46 am
That's is correct.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message