Literature on crossclassified with bayes PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Bèiendé Niú posted on Tuesday, July 25, 2017 - 10:20 am
Dear MPlus Developers and Communitiy,

I am desperately looking for comprehensible and application-oriented literature on crossclassified analysis with the bayesean estimator. (i.e. not too technical, rather textbook-style)

This literature should help to answer questions such as:

Which assumptions (apart from the crosscalssified data structure) must be met in the data before such modelling is appropriate (are there even any? Distributions? Outliers? etc.)?

How do I intepret the outcomes correctly and are there any differences to the interpretations of coefficients in "frequentist" multi-level models?

I would be very grateful for any suggestion!

Thank you and kind regards
 Bengt O. Muthen posted on Tuesday, July 25, 2017 - 6:00 pm
I don't know if your situation has cross-classified data or longitudinal data. If the former, we just posted a Psychometrika paper yesterday under Papers, Bayesian Analysis and we have a paper on this too:

Asparouhov, T. & Muthén, B. (2015). General random effect latent variable modeling: Random subjects, items, contexts, and parameters. In Harring, J. R., & Stapleton, L. M., & Beretvas, S. N. (Eds.), Advances in multilevel modeling for educational research: Addressing practical issues found in real-world applications. Charlotte, NC: Information Age Publishing, Inc. Mplus scripts.

None of these papers can be classified as application oriented - they are technical but have applications.

If your situation involves longitudinal data - see our upcoming Johns Hopkins workshop on DSEM and also our new web page

http://www.statmodel.com/TimeSeries.shtml

The usual assumption we use is that the random effects have normal distributions. The observed variables can however be binary or ordinal as well.

Interpretations are very similar comparing ML and Bayes. See chapter 9 of our new book Regression and Mediation Analysis Using Mplus.
 Bèiendé Niú posted on Thursday, August 03, 2017 - 4:35 am
Thank you very much for the helpful suggestions.

One mre thing: My question on "Distributions" was anwerd only by mentionning the scale of measurement of the data.

All my data is interval/ratio scaled, but I am worried about non-normal distributions, in particular skewness.

Is that an issue with the Bayesian Estimator in cross-classified multi-level analysis? (only HLM, not MSEM)
 Bengt O. Muthen posted on Thursday, August 03, 2017 - 5:26 pm
Usually, results are rather robust to non-normality of continuous variables as long as they don't show strong floor or ceiling effects. It is a research question how much that affects Bayesian analysis which assumes normality. You can do simulations in Mplus to study it.
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: