The DHS Program User Forum: Merging data files

Home » Data » Merging data files » Merging Datasets PDHS

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

Merging Datasets PDHS [message #25425]

Tue, 18 October 2022 07:26

adeen97
Messages: 3
Registered: October 2022

Member

Hello, I am trying to see the impact of the Benazir Income Support Programme (BISP) -which is an unconditional cash transfer program initiated by the government of Pakistan- on the utilization of reproductive health services (antenatal, postnatal care, skilled assistance during delivery, and place of delivery). However, I am a bit confused about how to merge the files

Maternal health variables and the question of whether the individual is a BISP recipient are present in the IR file, but these variables are also present in the BR and KR files (the BISP variable is also present in the CR file)

Can someone please guide how I should go about merging? I also need to merge the HR and PR files as I plan on creating a poverty scorecard, similar to the one created by the program using household socio-economic indicators to assess program eligibility.

Report message to a moderator

Re: Merging Datasets PDHS [message #25427 is a reply to message #25425]

Tue, 18 October 2022 08:29

Bridgette-DHS
Messages: 3188
Registered: February 2013

Senior Member

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

You probably do not need to do these merges. The cases in the IR file are the women respondents. The cases in the KR and BR files are the children of those women, with one record for each child. The mother's information (including this variable) is included on the child's record. If you did a merge, you would just be duplicating information that is already there.

The cases in the HR file are households. The cases in the PR file are the household members, with one record for each household member. The HR and PR files have the same information, just arranged differently.

Report message to a moderator

Re: Merging Datasets PDHS [message #25429 is a reply to message #25427]

Tue, 18 October 2022 09:39

adeen97
Messages: 3
Registered: October 2022

Member

Thank you so much for your response! so basically the child file (KR) is a better option given it has information on the maternal variables as well as the program variable.

Report message to a moderator

Re: Merging Datasets PDHS [message #25431 is a reply to message #25429]

Tue, 18 October 2022 13:14

Bridgette-DHS
Messages: 3188
Registered: February 2013

Senior Member

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

Right--the BR file has more children, but the KR file, with children born in the past five years, has the variables that describe child health, etc., and would be much more useful.

Report message to a moderator

Re: Merging Datasets PDHS [message #25456 is a reply to message #25431]

Mon, 24 October 2022 11:38

adeen97
Messages: 3
Registered: October 2022

Member

Thank you so much for your response! one last question, as I will be now using the KR file, I also need to assess certain household characteristics (highest education level of HH head, toilet facility, ownership of refrigerator, A/C, car, livestock, agricultural land, etc) so should I merge with the PR file or HR?

Report message to a moderator

Re: Merging Datasets PDHS [message #25467 is a reply to message #25456]

Tue, 25 October 2022 08:42

Bridgette-DHS
Messages: 3188
Registered: February 2013

Senior Member

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

I suggest that you use the PR file, but just keep the line for the household head (the household member with hv101=1), since you have a special interest in that person's characteristics. The following Stata lines show how to do that. Good luck!

* specify workspace
cd e:\DHS\DHS_data\scratch

use "...PKPR71FL.DTA", clear 
keep if hv101==1
gen cluster=hv001
gen hh=hv002
sort cluster hh
save PKtemp.dta, replace


use "...PKKR71FL.DTA", clear 
gen cluster=v001
gen hh=v002
sort cluster hh
merge cluster hh using PKtemp.dta

tab _merge

* _merge is 2 for households that do not have a child in the KR file
keep if _merge==3
drop _merge

Report message to a moderator

Previous Topic:	Merging KR with PR
Next Topic:	Obtaining BMI data for males in MR dataset DHS 2019-2021: NFHS5

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Thu Nov 7 11:19:48 Coordinated Universal Time 2024