| Re: Missing values for anemia [message #9312 is a reply to message #9294] |
Wed, 09 March 2016 06:42   |
clarapd
Messages: 8
Registered: November 2015
|
Member |
|
|
Thanks for you help. I did not understand this because some of these questions are only asked to subsamples of the populations and this is country specific and it's not explained in the questionnaires.
Now, knowing this, I'm performing the analysis for 29 different countries. When I perform the analysis at a country level, I undestand that I need to use the individual weights even if only a subsample answered the questions. My doubt is: when I append all datasets (29 countries) I need to rescale the weights to make each country representative of their population size, is that right? And for the rescale proces, I think that I need to drop (or to not take into account) the (.) values to make the subsample representative of the whole country population.
Thanks you very much.
Clara Pons Duran
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
