Hi everybody
I am doing a latent class analysis with gsem. Additionally, I'm using syv:, as I have both weights and strata that I want to account for. However, STATA does not allow me to to run the following command for goodness-of-fit:
It gives the error: "invalid subcommand lcgof".
Without svy: and with [pweight=pervgt] I can run the command. I have also tried running a separat regressions with, respectively, svy: and with [pweight=pervgt] and the results are the same. So, with this in mind, could I safely ignore that I have not accounted for strata in my latent class analysis?
I am doing a latent class analysis with gsem. Additionally, I'm using syv:, as I have both weights and strata that I want to account for. However, STATA does not allow me to to run the following command for goodness-of-fit:
Code:
svy: gsem (var1 var2 var3 var4 var5 var6 var7 var8 var9 var10 var11 var12 var13 var14 var15 <- ) if int_resultat == 1, logit lclass(C 2) nonrtolerance startvalues(randomid, draws(15) seed(123321)) estat lcgof
Without svy: and with [pweight=pervgt] I can run the command. I have also tried running a separat regressions with, respectively, svy: and with [pweight=pervgt] and the results are the same. So, with this in mind, could I safely ignore that I have not accounted for strata in my latent class analysis?
