|
|
Multilevel or multigroup analysis |
|
Message/Author |
|
Qiao Hu posted on Thursday, September 08, 2011 - 8:37 am
|
|
|
I use the data from five countries to do a latent class analysis, could I do a multilevel analysis as well (i.e.individual level and country level )? |
|
|
I would not recommend that with such few countries. Instead of treating country as a random mode of variation, I would treat it as fixed. That is, I would include dummy covariates to represent countries, or use multiple-group analysis. |
|
Back to top |
|
|