|
|
Estimating prevalence of classes from... |
|
Message/Author |
|
|
Hello , I run LTA analyses on the complex survey data. I use the following syntax: classes = c1(7) c2(7) c3(7) c4(7); strat=varstrat; cluster=varpsu; weight=weight; Analysis: Type=complex mixture; starts=30 5; algorithm=EM; However, in the output file FINAL CLASS COUNTS AND PROPORTIONS FOR EACH LATENT CLASS VARIABLE BASED ON THE ESTIMATED MODEL provide unweighted estimates. Are item response and transition probabilities found in the output file are weighted? Can you please explain where I can find weighted estimates? I truly appreciate your help with this. |
|
|
All the estimates are weighted. If you think your estimates are not weighted send your example to support@statmodel.com. |
|
Back to top |
|
|