Pakistan 2017-2018 Excluded Territories [message #29693] |
Tue, 23 July 2024 02:53 |
UnseenDisability
Messages: 14 Registered: February 2024
|
Member |
|
|
Dear DHS,
I am confused about the exclusion of Azad Jammu and Kashmir as well as Gilgit Baltistan. My understanding is that they are included in survey data, but not in national statistics for political reasons. When analyzing survey data, is their data a normal part of the included data or must they be separately included? How could I weight survey data separately, or exclude results for these areas from analysis?
Thank you in advance for any advice.
|
|
|
Re: Pakistan 2017-2018 Excluded Territories [message #29707 is a reply to message #29693] |
Wed, 24 July 2024 08:26 |
Janet-DHS
Messages: 878 Registered: April 2022
|
Senior Member |
|
|
Following is a response from DHS staff member, Tom Pullum:
As you say, these areas are excluded from national estimates for political reasons. DHS has an agreement with the government of Pakistan that they should always be analyzed separately.
The usual weight variable, hv005 or v005, is coded 0 in those areas. You will see this if, in Stata, you open the PR file (for example) and enter "tab hv024, summarize(hv005)". This means that in the usual tabulations, or in any weighted analyses that you do, these areas will be dropped automatically. You do not need to do anything to exclude them.
To analyze these areas separately, use shv005 or sv005. This weight is 0 everywhere EXCEPT int these areas.
If you try to make a variable that combines hv005 and shv005, the balance will be seriously incorrect.
|
|
|
|