The DHS Program User Forum: Weighting data » Survey weights for an appended file

Home » Data » Weighting data » Survey weights for an appended file

Show: Today's Messages :: Show Polls :: Message Navigator

Survey weights for an appended file [message #4098]

Tue, 31 March 2015 07:36

rejone4@emory.edu
Messages: 8
Registered: November 2014
Location: United States

Member

I am appending 6 different country files together to do some pooled analysis.
I have looked through previous answers about how to denormalize the data and then weight the different files prior to append which I understand. However, should I make any additional adjustments once I have the pooled file for the overall population and sum up across countries?

Report message to a moderator

Re: Survey weights for an appended file [message #4107 is a reply to message #4098]

Wed, 01 April 2015 15:33

Trevor-DHS
Messages: 808
Registered: January 2013

Senior Member

You should continue to use your newly calculated weights from the denormalization process, and continue to use svy commands (or the equivalent if you are using other software than Stata) using strata variables that are coded for each separate strata in your pooled dataset (that is don't have strata with the same code from different surveys - that also applies when pooling two surveys in the same country). That hopefully should be enough.

Report message to a moderator

Previous Topic:	Weighting after de-normalization
Next Topic:	De-normalizing and weighting data for multiple countries

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Thu Jan 8 06:42:00 Coordinated Universal Time 2026