[Limdep Nlogit List] correcting for survey data design in discrete choice models

[Patricia Duffy]

I am working survey data -- cross-sectional -- that was has
stratification and
clusters (e.g. psu's), with variables that can be used to identify
stratification and cluster.

I have been trying to find in LIMDEP how to correct for these survey 
design effects in the variance estimation in logistic models (binary and
ordered).

I would greatly appreciate any help with setting up the LIMDEP code to
account for survey design.  

I have NLOGIT 3 / LIMDEP 8.

Thanks for any answers.

Patricia Duffy
Auburn University