I am doing a simulation study for growth mixture model (Please see the below my code) with 100 replications. After running my code, Mplus provides one output file including parameter estimates and other information indicies. I think this is a result from first replication. Am I correct?
In that output automatically provided by mplus, I am wondering that the "FINAL CLASS COUNTS AND PROPORTIONS FOR THE LATENT CLASS PATTERNS BASED ON ESTIMATED POSTERIOR PROBABILITIES" is the result of class proportions based on first generated data?
If so, the result was very different from one I generated. For example, I generated two latent classes which class proportions for class 1 and 2 are 80% and 20% respectively, but the output provided by mplus showed that
1 318.28993 0.53048 2 281.71007 0.46952
However, I ran a mixture analysis using the first generated data and used the same mplus code under the analysis (but in that code, I don't include the monte carlo and model population). Then, I got the results I want. They were very close to 80% ans 20%.