What about Missing Data?
Missing data in the latent class indicators is acceptable and will be addressed via maximum likelihood estimation in the LCA or LTA models. The model assumes that data are missing at random and it may be worth checking this assumption prior to the analysis.
Listwise deletion will be used for cases missing data on covariates, however. At present, there is no simple way to use multiple imputation in LCA or LTA, although this is an area of active research and development at the Penn State Methodology Center (http://methodology.psu.edu/). As such, we were not able to retain cases missing data on any of our covariates in our conditional RMLCA models.
RMLCA can be conducted in MPLUS or SAS proc lca and should be formatted according to the program you will be using. We ran the same models in both MPLUS and SAS to ensure convergent results.