I ran a longitudinal mixture model (c=5) over several measurement waves of count data. The output of this analysis does not give the mean growth factors for each latent class as shown in the graph (plot 2).
How can I obtain the mean intercept and slope factors for each latent class, as well as their significance level?
I found the growth factor means, but these are the log-transformed values (as I use count-data). For instance the intercepts are negative, whereas the scale ranges from 0 to 4. Is there a means of obtaining the "real" values that match the graphs?
I think you will find what you want if you ask for RESIDUAL in the OUTPUT command. These are the values that are plotted.
socrates posted on Thursday, February 26, 2009 - 3:39 am
When running a GMM with two parallel processes (one continuous and one count variable) and adding a variable to predict growth parameter variances of the continuous variable, I always get the following error message:
THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO AN ILL-CONDITIONED FISHER INFORMATION MATRIX. CHANGE YOUR MODEL AND/OR STARTING VALUES.
THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO A NON-POSITIVE DEFINITE FISHER INFORMATION MATRIX. THIS MAY BE DUE TO THE STARTING VALUES BUT MAY ALSO BE AN INDICATION OF MODEL NONIDENTIFICATION. THE CONDITION NUMBER IS 0.442D-10.
THE STANDARD ERRORS OF THE MODEL PARAMETER ESTIMATES COULD NOT BE COMPUTED. THIS IS OFTEN DUE TO THE STARTING VALUES BUT MAY ALSO BE AN INDICATION OF MODEL NONIDENTIFICATION. CHANGE YOUR MODEL AND/OR STARTING VALUES. PROBLEM INVOLVING PARAMETER 64.
Parameter 64 refers to the last measurement occasion of the continuous variable in THETA.
Do you have any idea why this happens and how to solve this problem?