The DHS Program User Forum: Other countries » regions AJK and GB in PDHS (Pakistan DHS)

Home » Countries » Other countries » regions AJK and GB in PDHS (Pakistan DHS)

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

regions AJK and GB in PDHS (Pakistan DHS) [message #24819]

Fri, 15 July 2022 22:43

chande
Messages: 12
Registered: March 2021

Member

I am looking for the impact of WASH on child malnutrition.
Is it possible to include Azad Jammu & Kashmir and Gilgit Baltistan regions from PDHS in my analysis?

Report message to a moderator

Re: regions AJK and GB in PDHS (Pakistan DHS) [message #24850 is a reply to message #24819]

Wed, 20 July 2022 16:36

Janet-DHS
Messages: 938
Registered: April 2022

Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

Those two regions can be analyzed separately (an analysis for AJK and a separate analysis for GB) but you cannot pool them and you cannot pool them with the rest of Pakistan. In the IR file, for example, the usual sample weight, v005, is 0 for those regions. The correct weight variable for these two regions is sv005, and it is normalized to have a mean of 1 (or 1000000) in each of them. To repeat, you cannot pool these regions with each other or with the rest of the country. The government of Pakistan has been very clear about not wanting users to do that.

Report message to a moderator

Re: regions AJK and GB in PDHS (Pakistan DHS) [message #27709 is a reply to message #24850]

Sun, 24 September 2023 23:06

owraza
Messages: 37
Registered: December 2013
Location: Nebraska, US

Member

What is the reason for not pooling these two regions? I do understand that pooling them in rest of Pakistan and doing comparison analysis with previous PDHS would be inappropriate, but what if that is not the objective? What if I'm interested in using only PDHS 17-18. Also, related to the same issue, how appropriate it would be to AJK and GB regions in geographically weighted regression analysis in which spatial weights are used and would I still need to use sampling weights?

Report message to a moderator

Re: regions AJK and GB in PDHS (Pakistan DHS) [message #27732 is a reply to message #27709]

Wed, 27 September 2023 09:28

Janet-DHS
Messages: 938
Registered: April 2022

Senior Member

Following is a response from DHS staff member, Tom Pullum:

The surveys are "owned" by the countries. The government of Pakistan, for political reasons, did not want to include AJK and GB in the national estimates, because these are disputed territories. We at DHS must respect that decision. If you want to include AJK and GB in a national estimate, you can do that, of course, but we will not facilitate that.

Report message to a moderator

Re: regions AJK and GB in PDHS (Pakistan DHS) [message #27744 is a reply to message #27732]

Wed, 27 September 2023 22:35

owraza
Messages: 37
Registered: December 2013
Location: Nebraska, US

Member

Thanks.

Report message to a moderator

Previous Topic:	Need ethical review documents
Next Topic:	Pakistan DHS 2012 district codes

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sun Oct 12 13:37:44 Coordinated Universal Time 2025