The DHS Program User Forum: Wealth Index » Wealth quintile for Sub-sample

Home » Topics » Wealth Index » Wealth quintile for Sub-sample

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

Wealth quintile for Sub-sample [message #27947]

Tue, 24 October 2023 21:41

Rupon
Messages: 16
Registered: October 2023

Member

Dear all,
We are running regression analysis of health insurance. As a sensitivity check, we are considering random sub-samples (40, 60, 80 percent) from the full sample. Since the wealth quintiles are relative to the population, is it necessary that we redefine the wealth quintiles for every sub-sample we draw? Similarly, is it recommended that we do the same for sub-population analysis such as by caste, residence? Moreover, can we use wealth index(rural/urban) instead of combined wealth index?
Looking forward to your suggestions.

Thank you
Regards
Rupon

Report message to a moderator

Re: Wealth quintile for Sub-sample [message #27951 is a reply to message #27947]

Wed, 25 October 2023 10:33

Janet-DHS
Messages: 938
Registered: April 2022

Senior Member

Thank you for submitting your question. Could you please provide some more detailed information so we can better advise you?

Can you let us know:
• Which survey you are using (Include country name and year)?
• Which data files you are referring to?
• Which software you are using (Stata, SPSS, R, etc.)?

If you are trying to match a Table in a final report, please also indicate which table and which estimate you are trying to match.

Report message to a moderator

Re: Wealth quintile for Sub-sample [message #27952 is a reply to message #27951]

Wed, 25 October 2023 10:44

Rupon
Messages: 16
Registered: October 2023

Member

Dear Janet,
Below I provide the details as asked:
Country: India, NFHS-5 2019-21 survey
Data: NFHS - Men and Women Recode files (appended)
Software used: Stata

Thank you

Report message to a moderator

Re: Wealth quintile for Sub-sample [message #27955 is a reply to message #27947]

Wed, 25 October 2023 13:01

Janet-DHS
Messages: 938
Registered: April 2022

Senior Member

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

I recommend that you do NOT re-calculate the wealth index or quintiles. I don't see any reason for doing that. I do see several reasons for NOT doing it, including the fact that no one else could replicate your results, because you are drawing random samples and there would be differences in different samples.

If your analysis treats the urban and rural populations as completely different populations, then you can use the urban wealth index in the urban sample and the rural wealth index in the rural sample. However, if you are pooling the urban and rural cases, which is more common, then you do not want to use the urban wealth index in the urban part of the sample and the rural wealth index in the rural part of the sample. Instead, use the combined wealth index.

Report message to a moderator

Re: Wealth quintile for Sub-sample [message #27958 is a reply to message #27955]

Wed, 25 October 2023 14:29

Rupon
Messages: 16
Registered: October 2023

Member

Thank you so much.

Report message to a moderator

Previous Topic:	Lesotho Urban/Rural Ns
Next Topic:	Tabulating urban/rural residence by wealth quintiles

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Fri Oct 31 16:12:50 Coordinated Universal Time 2025