Minimum number of groups for conducting multilevel latent class analysis

[Yong-Jin Ahn]

Hi Statalisters,

My question is as the title. I am working on a survey data that has around 9,000 observations (level-1) with 24 identifiable groups (level-2). I haven't dived into the analysis yet, but seems a little too small a number of groups to conduct MLCA.
I know Stata currently does not support MLCA, so planning to use another software (e.g., Mplus), but thought I might seek some advice to save some time. Anyone has ideas or recommended references in this area? Thanks in advance.

[Yong-Jin Ahn]

Just found a reference that I think is relevant.
Park J, Yu HT. Recommendations on the Sample Sizes for Multilevel Latent Class Models. Educ Psychol Meas. 2018;78(5):737-761. doi:10.1177/0013164417719111