It sounds as if you are using the wrong weights. Not sure what is your sampling design. For example, if you are using a two-stage clustered sampling in which the PSUs are selected with replacement with PPS (say, p(i)) and the SSU are selected SRS without replacement, your weights should look something like this:
pw=1/n*1/p(i)*M(i)/m(i)
where n is the number of PSUs sampled,
M(i) is the number of SSUs in PSU(i), and
m(i) is the number of SSUs sampled in PSU(i).
And you should use the second svyset syntax, not the first one, which is appropriate for a single-stage stratified random sample.
Finally, the population size reported by STATA corresponds to the number of SSUs, not PSUs.
Hope that helps.
Albert.
________________________________________
From: [email protected] [[email protected]] On Behalf Of Tomeka Davis [[email protected]]
Sent: Monday, November 09, 2009 4:33 PM
To: [email protected]
Subject: st: help interpreting SVY output
Hello -
This is my first time using the set of SVY commands in Stata and I have a few questions. When I use the syntax:
svyset psu [pweight=pw], strata (strata)
Stata 1) does not calculate the standard errors and 2) I get a message stating that standard errors are" missing because of stratum with single sampling units".
Furthermore, I am also getting incorrect counts for my PSU. There were originally only 100 PSUs in my data. However, I do get standard errors with the following syntax:
svyset psu [pweight=pw], strata (strata) || _n, singleunit(certainty)
But I still get a 455 PSU count. Again, I have never used SVY to account for clustering and complex survey design. I would appreciate any advice.
Thank you.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
