The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Wealth Index » Rural wealth index construction
Rural wealth index construction Tue, 23 February 2016 14:34
 mmr-UMICH Messages: 21Registered: February 2015 Location: A2, MI Member
Hi DHS Forum

I wanted to calculate the rural wealth index for Bangladesh DHS 2011 survey. Downloaded an SPSS syntax file from DHS website which used QH series variables that is, SPSS code uses DHS raw data. We have access to only Recode data, which has HV series variables.

So I was trying to convert existing SPSS code for Recode data, but I could not figure out the corresponding HV variables for the following QH variables:
-----------------------------------------
qh110A qh110B qh110C qh110D qh110E qh110F qh110G qh110H qh110I qh110J qh110K qh110L
qh118A qh118B qh118C qh118D qh121 qh122a qh122b qh122c
qh122_1 to qh122_5 qh123 and DOMESTIC.

Thank you so much.

The portion of the original code is given below:
*{Reset missing values to "does not have", change 2 code to 0}.
if (missing(qh110a) | qh110a<>1) qh110a=0.
if (missing(qh110b) | qh110b<>1) qh110b=0.
if (missing(qh110c) | qh110c<>1) qh110c=0.
if (missing(qh110d) | qh110d<>1) qh110d=0.
if (missing(qh110e) | qh110e<>1) qh110e=0.
if (missing(qh110f) | qh110f<>1) qh110f=0.
if (missing(qh110g) | qh110g<>1) qh110g=0.
if (missing(qh110h) | qh110h<>1) qh110h=0.
if (missing(qh110i) | qh110i<>1) qh110i=0.
if (missing(qh110j) | qh110j<>1) qh110j=0.
if (missing(qh110k) | qh110k<>1) qh110k=0.
if (missing(qh110l) | qh110l<>1) qh110l=0.
if (missing(qh118a) | qh118a<>1) qh118a=0.
if (missing(qh118b) | qh118b<>1) qh118b=0.
if (missing(qh118c) | qh118c<>1) qh118c=0.
if (missing(qh118d) | qh118d<>1) qh118d=0.
if (not(missing(qh122c)) & qh122c < 99.98) landarea=qh122c.
if (qh122c=99.95) landarea=95.
if (missing(qh122b) | qh122b<>1) landarea=0.
frequencies landarea.
if (missing(qh121) | qh121 <>1) qh121=0.
if (missing(qh122_1) | qh121 <>1) qh122_1=0.
if (missing(qh122_2) | qh121 <>1) qh122_2=0.
*if (missing(qh122_3) | qh121 <>1) qh122_3=0.
if (missing(qh122_4) | qh121 <>1) qh122_4=0.
if (missing(qh122_5) | qh121 <>1) qh122_5=0.
if (missing(qh122a) | qh122a<>1) qh122a=0.
missing values qh122_1 to qh122_5 (98,99).
if (missing(qh123) | qh123<>1) qh123=0.
Re: Rural wealth index construction [message #9236 is a reply to message #9205] Thu, 25 February 2016 18:32
 Liz-DHS Messages: 1516Registered: February 2013 Senior Member
Dear User,
Does the attachment help?
These are the CWI coefficients for Bangladesh. If you still need assistance, please let us know.
Thanks!
Re: Rural wealth index construction [message #9270 is a reply to message #9236] Tue, 01 March 2016 14:22
 mmr-UMICH Messages: 21Registered: February 2015 Location: A2, MI Member
Thank you, Liz.

Re: Rural wealth index construction [message #11465 is a reply to message #9236] Wed, 28 December 2016 16:47
 r01efa16 Messages: 10Registered: November 2016 Location: UK Member
Hello Liz,

I am trying to construct Wealth Index for Egypt using STATA.
I refered to the manuals (attached here) in regard of how to calculate wealth index, however all using SPSS syntax, So I am a bit struggling

Have you done the coding using STATA?
Could you tell me how to score the PCA components based on the regression method?

What weight should do with the index? How can I use weight after constructing the index and divide the quintiles?