[Limdep Nlogit List] Maximum Number of Variables in a Latent Class Model
Sebastian Feld
sfeld at uni-muenster.de
Thu Oct 30 03:30:00 EST 2008
Hello everyone,
I am currently estimating a Latent Class Poisson Regression Model, in which
I identified two latent classes according to the BIC criterion.
I am experiencing some problems with that model. I specified an intercept
term and 7 predictor variables within the 2 segments. After that I tried to
profile the segments with heterogeneity covariates using the lcm= command.
As long as I specify a maximum number of four heterogeneity covariates the
model works fine. But when I specify an additional covariate (e. g. 7
predictors and 5 covariates) the program shuts down everytime.
So my question is if there is a maximum number of predictor an/or covariate
variables in a Latent Class Model in Limdep? If that is the case is there
any possibility to incorporate more covariates?
I appreciate your help.
Best, Sebastian
More information about the Limdep
mailing list