I posted a solution that requires the SAS IML package.
Regards,
--Alex Cavallo
Lexecon
Statalisters-the message below was posted to the SAS
list serve, but no one could provide an answer. Is
anyone aware of the SAS command or option that will
correct for clustered observations in the same manner
as the "cluster" option in Stata? Tim
(Original message)
----------------------------------------------------
Hi SAS-L,
I'm trying to obtain standard errors for a simple
linear regession that
are corrected for intra-cluster residual correlations
and for heteroskedasticity a la Huber/White. In other
words, I'd like to replicate the standard errors
produced by STATA when using the post-
estimation options ", cluster(<clustervar>) robust".
I've seen a suggestion for logit specifications to use
the "repeated" option in proc genmod and then to
multiply the standard errors obtained by
the square root of the number of clusters divided by
the number of clusters minus 1 to match STATA's
results.
Is this the best way to proceed? I'm also wondering
if this is the most efficient way to proceed as I'm
using a data set with about 10 million
observations and a large number of variables.
Thanks kindly for any help that you can offer.
Grant Miller
=====
[email protected]
*
* 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/
