The DHS Program User Forum: Wealth Index » Quintile cutoffs

Home » Topics » Wealth Index » Quintile cutoffs - weighted and unweighted

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

Quintile cutoffs - weighted and unweighted [message #9804]

Sun, 22 May 2016 20:44

KenzoFry
Messages: 21
Registered: December 2015
Location: UK

Member

Hi,

It appears that in the composite tab of many of the the excel files, the quintile cut-offs are given from an unweighted analysis. For example, Kenya DHS 2014 or Cambodia DHS 2014 have an n that is the same as the number of respondents in the dataset. In other countries, such as Philippines DHS 2013 or Madagascar DHS 2008 the quintile cutoffs are presented from a weighted analysis. It would be great to know why this is. Also, am I right to think that generally, the quintile cutoffs should be calculated after weighting using hhmemwt?

Thanks

Kenzo

EquityTool.org

Report message to a moderator

Re: Quintile cutoffs - weighted and unweighted [message #10107 is a reply to message #9804]

Tue, 28 June 2016 10:18

k322a
Messages: 3
Registered: June 2016

Member

Hi,

Hope to know if you managed to solve this? I am also having issues replicating the quintile (cutoffs). This is my code for the 2010 Tanzania DHS -

xtile _5=hv271 [pw=cond(hv012==0,hv005*hv013,hv005*hv012)] , n(5)

. ta _5 hv270

5 |
quantiles | Wealth index
of hv271 | Poorest Poorer Middle Richer Richest | Total
-----------+------------------------------------------------ -------+----------
1 | 1,889 93 0 0 0 | 1,982
2 | 0 1,802 12 0 0 | 1,814
3 | 0 0 1,711 0 0 | 1,711
4 | 0 0 188 1,841 0 | 2,029
5 | 0 0 0 221 1,866 | 2,087
-----------+------------------------------------------------ -------+----------
Total | 1,889 1,895 1,911 2,062 1,866 | 9,623

Any help would be much appreciated.

Kerry

Report message to a moderator