Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: standard deviation (not se) for survey data

From   Laura Piersol <>
Subject   st: standard deviation (not se) for survey data
Date   Wed, 8 Sep 2004 17:57:42 -0700 (PDT)

Dear Stata-Listers, 

I have survey data (including strata, weights, and clusters) and I am trying to
estimate the standard deviation (not standard error) of the population. I know
Stata estimates the standard errors with a linearization for this type of
problem, so maybe Stata doesn't calculate the SD of the population as it would
under SRS??

The most relevant Stata reference I could find suggested backing out the SD
from “deff”. Since the denominator of “deff” (resulting from the command
svymean) is the variance of the mean under simple random sampling, multiplying
that denominator by n will return sigma^2. This is a way to get the population
standard deviation under simple random sampling, but this will not take account
of the sampling design as far as I can tell.

Does Stata calculate the population standard deviation which accounts for the
sampling design? 

Many many thanks for any help,


Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
*   For searches and help try:

© Copyright 1996–2021 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index