| Re: Wealth Index data disaggregation [message #25920 is a reply to message #25906] |
Tue, 03 January 2023 20:31  |
Bridgette-DHS
Messages: 3199
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
All the wealth quintiles are constructed with the HR file and the continuous index, hv271. The urban and rural versions are just constructed separately for the urban households and the rural households. The following code will match DHS exactly, except for perhaps a couple of households that are involved in ties when the continuous values are sorted.
Note that you have to copy the constructed values (in the HR file) to the PR and other files, with merges using just the cluster and household codes (and region for an India survey). Everyone (woman, man, and child) in the same household has the same value of the continuous wealth index and the quintiles.
The program is set up for a survey that already includes the urban and rural quintiles, so that you can check the calculation. You would have to change the program slightly for a survey that actually does NOT include the urban and rural quintiles.
* SIMPLE STATA CODE TO CONSTRUCT WEALTH QUINTILES THAT ALMOST MATCH DHS
set more off
use "...\HTHR71FL.DTA" , clear
keep hv001 hv002 hv005 hv012 hv013 hv025 hv270* hv271*
* calculate overall quintiles
gen mem = hv012
replace mem = hv013 if mem == 0
gen pwt=mem*hv005
gen wt=pwt/1000000
xtile hv270_test=hv271 [pweight=pwt], nquantiles(5)
tab hv270 hv270_test
* calculate urban quintiles
xtile hv270_test1=hv271 if hv025==1 [pweight=pwt], nquantiles(5)
tab hv270a hv270_test1 if hv025==1
* calculate rural quintiles
xtile hv270_test2=hv271 if hv025==2 [pweight=pwt], nquantiles(5)
tab hv270a hv270_test2 if hv025==2
gen hv270a_test=.
replace hv270a_test=hv270_test1 if hv025==1
replace hv270a_test=hv270_test2 if hv025==2
correlate hv270a hv270a_test
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
