Date: Tue, 3 Jan 2006 08:44:30 -0600
Reply-To: Warren Schlechte <Warren.Schlechte@TPWD.STATE.TX.US>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Warren Schlechte <Warren.Schlechte@TPWD.STATE.TX.US>
Subject: Re: Post-Stratification Variance
Content-Type: text/plain; charset="us-ascii"
While I knew about how stratification does not always improve precision
over SRS, I was surprised at how a post-stratified sample reduced the
error in all strata. In pre-stratified samples, strata are typically
thought of as being independent, hence total variance is obtained by
summing the strata variances (no covariances).
What seems odd is that in post-stratification, the samples, because of
the ad-hoc way of creating the strata, are no longer independent.
However, the variance of the total is still computed using the sum of
the variances, even as the individual strata variances are related, for
all are computed using the same "n" and "N".
I really appreciate your input, and must think about this further.
HOH Fisheries Science Center
5103 Junction Hwy
Ingram, TX 78025
Phone 830.866.3356 x.214