Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Need help on variance estimation using replication methods while incorporating raking


From   Steven Samuels <sjsamuels@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Need help on variance estimation using replication methods while incorporating raking
Date   Tue, 22 Nov 2011 19:22:28 -0500


Several of your beliefs appear erroneous to me, because:

1. Raking is intended to decrease bias, not variance. In fact it will increase variance slightly because the raked weights tend to be more variable than the originals.

2. Weight trimming is intended to decrease variance at the expense of an increase in bias.  Thus the overall goal is to decrease mean square error (bias^2 + variance^2).

In Stata, you can create bootstrap replicates with Stas Kolenikov's -bsweights- module (-findit-).

To quote from the -help-

 "Note that for proper results, the survey agency must provide the original probability weights as inputs to bsweights. The same adjustment procedure that produces the publicly available weights from those probability weights should be applied to the bootstrap weights."

In other words, you must repeat the raking procedure for each replicate. 

I suggest that you use Stata's default linearization variance method on your raked data. This is standard practice for those surveys, the majority, which do not provide replicates. The standard errors will not reflect  added variability due to estimating the weights, true, but this increase will usually be offset by other factors that make the estimated standard errors conservative. 


Steve





On Nov 22, 2011, at 5:53 PM, Bilal Khan wrote:

Hi

I am using stata 10 and 11.

I  have a multistage or complex survey data which I raked using some auxiliary variables like previous votes in different elections (highly correlated with output variables). Now I want to find 95 percent confidence intervals for my estimated and I believe I can use svy option in stata to calculate such estimates through Taylors series. However, this would not incorporate raking into design which perhaps may lower the sampling error. I can trim weights as well but I would lose precision of estimates.

So I plan to use replication methods which I believe can cater for raking or post stratification. I do have the commands to do so in stata but I am not sure how to create replicate weights before using replication methods for survey variance estimates. Can anyone suggest an easy way to create replication weights and incorporate replicated weight; especially bootstrap in variance estimation. Also raking tends to decrease sampling variance. Would this be reflected in the variance estimation using replication methods like bootstrap or Jackknife?

Thank you in advance for your help.
*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index