st: large elasticities in the latent class model

["Simona Rasciute" <S.Rasciute@lboro.ac.uk>]

Hello,

I would appreciate your comments on the following issue. 
When I estimate the latent class model with three classes 
or more, the elasticities (both mean and standard 
deviation) are extremely large. However, neither the 
estimated parameters nor their standard errors are large. 
Does it mean that the large elasticities reveal the 
problem that is not revealed by the estimated coefficients 
are their standard errors? According to the information 
criteria, 5 latent class should be chosen and everything 
seems fine until the elasticities are estimated. In order 
to avoid huge elasticities, I would have to use only two 
classes, which is not supported by information criteria. 
How would you tackle this problem?

Thanks a lot in advance for your help.

Kind Regards,
Simona
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/