Logistic MLM PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Sherry Leviner posted on Sunday, August 17, 2014 - 12:35 pm
I am interested in running a logistic MLM with complex survey data. The data have weights at Level-1 an Level-2. There is clustering and stratification of the data. I need to test random effects and cross-level interactions. Is this type of analysis possible in Mplus? If it is possible, which type of analysis should I use?
 Linda K. Muthen posted on Monday, August 18, 2014 - 10:50 am
You can do this using TYPE=TWOLEVEL RANDOM with the MLR estimator. See Example 9.2. You would add the CATEGORICAL, WEIGHT, BWEIGHT, and STRATIFICATION options.
 Jacqueline Sims posted on Thursday, December 15, 2016 - 8:06 am
Hello,

Can you please confirm that logistic MLM can be run with complex survey data, using the MLR estimator with survey weights? Based on some prior posts on this discussion board, I'm unclear as to whether utilizing the MLR estimator will give me probit regression coefficients from the default, or will in fact produce logistic regression coefficients.

Thanks very much.
 Bengt O. Muthen posted on Thursday, December 15, 2016 - 8:43 am
MLR can give either probit or logit with complex survey data including weights. Logit is the default link.
 Jacqueline Sims posted on Thursday, December 15, 2016 - 8:51 am
Thank you very much!
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: