The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Sampling » how to set svyset with three-stage sampling data (Nigeria 2013)
how to set svyset with three-stage sampling data (Nigeria 2013) [message #12728] Tue, 04 July 2017 12:50 Go to next message
machiyama is currently offline  machiyama
Messages: 6
Registered: April 2013
Location: UK
Member
Dear DHS Forum,

I started using the 2013 Nigeria DHS and have a question.

According to the description in the final report (see below), it looks the samples in the 2013 Nigeria DHS were selected in 3 stage, unlike the other DHSs.
I wonder if we could still use the same svyset command.

egen strata=group(sstate v025)
gen weight=v005/1000000
svyset v021 [pweight=weight], strata(strata)

How should we set svyset for this data?
Any advice will be appreciated.

Thank you, in advance.
Best regards,

-----------------------------------------------------
FYI- Here is description of the sampling in the final report.
Page 378 in the final report. https://dhsprogram.com/pubs/pdf/FR293/FR293.pdf

"The sample for the 2013 NDHS was a stratified sample, selected independently in three stages
from the sampling frame. Stratification was achieved by separating each state into urban and rural areas. In
the first stage, 893 localities were selected with probability proportional to size and with independent
selection in each sampling stratum.

In the second stage, one EA was randomly selected from most of the selected localities with an
equal probability selection. In a few larger localities, more than one EA was selected. In total, 904 EAs
were selected. After the selection of the EAs and before the main survey, a household listing operation was
carried out in all of the selected EAs. The household listing consisted of visiting each of the 904 selected
EAs, drawing a location map and a detailed sketch map, and recording on the household listing forms all
occupied residential households found in the EA with the address and the name of the head of the
household. If a selected EA included less than 80 households, a neighbouring EA from the selected locality
was added to the cluster and listed completely. The resulting list of households served as the sampling
frame for the selection of households in the third stage"


John Cleland, on behalf of the IUSSP Panel on Family Planning, Fertility and Urban Development
John.Cleland@lshtm.ac.uk
Re: how to set svyset with three-stage sampling data (Nigeria 2013) [message #12739 is a reply to message #12728] Thu, 06 July 2017 09:48 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3167
Registered: February 2013
Senior Member
Following is a response from Senior DHS Stata Specialist, Tom Pullum:


A more complex version of svyset would be desirable but we do not have enough information to construct it. The usual version is the best that is possible.


Re: how to set svyset with three-stage sampling data (Nigeria 2013) [message #12744 is a reply to message #12739] Thu, 06 July 2017 17:27 Go to previous messageGo to next message
machiyama is currently offline  machiyama
Messages: 6
Registered: April 2013
Location: UK
Member
Dear Bridgett and Tom,

Thank you very much for your reply. This is very helpful.



John Cleland, on behalf of the IUSSP Panel on Family Planning, Fertility and Urban Development
John.Cleland@lshtm.ac.uk
Re: how to set svyset with three-stage sampling data (Nigeria 2013) [message #13471 is a reply to message #12739] Tue, 07 November 2017 15:14 Go to previous messageGo to next message
claurti is currently offline  claurti
Messages: 2
Registered: September 2017
Member
Following up on the thread below ...

Isn't it possible to adjust for the last stage of selection (the selection of household within EA)? The Household Member Recode file has an identifier for Household.

Thank you,
Charles
Re: how to set svyset with three-stage sampling data (Nigeria 2013) [message #13475 is a reply to message #13471] Wed, 08 November 2017 08:24 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3167
Registered: February 2013
Senior Member

Following is another response from Senior DHS Stata Specialist, Tom Pullum:


The missing piece of information is the number of households in the EA (cluster) found during the household listing. We do not have that, and we do not have the number of households in the EA given in the sampling frame. (The probability of selection for the EA is proportional to that number.) This information is destroyed after the weights have been calculated, as part of DHS efforts to prevent identification of the EAS in the sample.

Previous Topic: Pooling male and female files
Next Topic: Rationale for eligibility criteria
Goto Forum:
  


Current Time: Mon Oct 7 05:05:25 Coordinated Universal Time 2024