The DHS Program User Forum      
Discussions regarding The DHS Program data and results
Home » Topics » Wealth Index » Comparative Wealth Index
Comparative Wealth Index [message #8868] Thu, 07 January 2016 05:16 Go to next message
nael is currently offline  nael
Messages: 6
Registered: January 2016
Member
Hi all,

I am studying how wealth affects child malnutrition in Rwanda with time. For that, I would like to calculate the Comparative Wealth Index for Rwanda for the years 1992, 2000, 2005 and 2010 (basically all available Rwanda DHS surveys).

Does anyone have the required Alpha and Beta values for these surveys?

Would greatly appreciate that as this information is vital for my PhD dissertation.

Thanks!
Re: Comparative Wealth Index [message #8869 is a reply to message #8868] Thu, 07 January 2016 10:18 Go to previous messageGo to next message
Liz-DHS
Messages: 1290
Registered: February 2013
Senior Member
Dear User,
Attached are the CWI coefficients.
Thank you for your post.
Re: Comparative Wealth Index [message #8871 is a reply to message #8869] Thu, 07 January 2016 12:17 Go to previous messageGo to next message
nael is currently offline  nael
Messages: 6
Registered: January 2016
Member
Thank you very much!!
Re: Comparative Wealth Index [message #8876 is a reply to message #8869] Sat, 09 January 2016 10:27 Go to previous messageGo to next message
nael is currently offline  nael
Messages: 6
Registered: January 2016
Member
Dear Liz-DHS,

Sorry to bother you again about the same issue. I used the values you sent me for Rwanda to calculate the CWI and compared the mean and std to the ones in table 3.1 in that following report http://dhsprogram.com/pubs/pdf/MR9/MR9.pdf.

Are the ones you sent me new updated alpha and beta scores or were they similar to the ones used in the report?

Specifically, I got this for 1992
Variable | Obs Mean Std. Dev. Min Max
-------------+---------------------------------------------- ----------
CWI | 6252 -1.00086 .3508722 -1.293119 1.913442
for 2000
Variable | Obs Mean Std. Dev. Min Max
-------------+---------------------------------------------- ----------
CWI | 9696 -1.176227 .7187736 -1.606931 3.206589
for 2005

Variable | Obs Mean Std. Dev. Min Max
-------------+---------------------------------------------- ----------
CWI | 10272 -9.14e-06 .5081679 -.2533463 4.451288
and for 2010
Variable | Obs Mean Std. Dev. Min Max
-------------+---------------------------------------------- ----------
CWI | 12540 -.0001001 .46742 -.787876 3.969633

whereas in table 3.1 they are different for all four cases.

Thank you again for your help!
Re: Comparative Wealth Index [message #8890 is a reply to message #8876] Tue, 12 January 2016 01:56 Go to previous messageGo to next message
Liz-DHS
Messages: 1290
Registered: February 2013
Senior Member
Dear User,
Here is a response from Dr. Shea Rutstein:
Quote:
The CWI is used to compare across surveys. There is no need to use it to compare regions. Since the transformation is constant within a survey, no difference in the concentration index should be observed (to the level of rounding).

If this does not answer your question, please post again.

[Updated on: Tue, 12 January 2016 11:13]

Report message to a moderator

Re: Comparative Wealth Index [message #8909 is a reply to message #8890] Wed, 13 January 2016 00:14 Go to previous messageGo to next message
nael is currently offline  nael
Messages: 6
Registered: January 2016
Member
Dear Liz-DHS,

Thank you for checking with the experts. However my question was why using the alpha and beta values you sent me in the excel sheet for Rwanda (1992, 2000, 2005 and 2010) gives different results for CWI than those in table 3.1 in the report in the following link http://www.dhsprogram.com/pubs/pdf/MR9/MR9.pdf for the same surveys.

For 2005 for example, using the values you sent me for alpha and beta, the mean for CWI is almost 0 whereas in table 3.1 the mean for CWI is -0.965 etc.. (standard deviations are also different as you can see from the tables I pasted in the previous message below, if you compared with table 3.1)

I hope I clarified my point, if not please let me know and I will attach files that detail the explanation more..

Thanks again for your assistance!
Re: Comparative Wealth Index [message #8910 is a reply to message #8890] Wed, 13 January 2016 00:15 Go to previous messageGo to next message
nael is currently offline  nael
Messages: 6
Registered: January 2016
Member
Dear Liz-DHS,

Thank you for checking with the experts. However my question was why using the alpha and beta values you sent me in the excel sheet for Rwanda (1992, 2000, 2005 and 2010) gives different results for CWI than those in table 3.1 in the report in the following link http://www.dhsprogram.com/pubs/pdf/MR9/MR9.pdf for the same surveys.

For 2005 for example, using the values you sent me for alpha and beta, the mean for CWI is almost 0 whereas in table 3.1 the mean for CWI is -0.965 etc.. (standard deviations are also different as you can see from the tables I pasted in the previous message below, if you compared with table 3.1)

I hope I clarified my point, if not please let me know and I will attach files that detail the explanation more..

Thanks again for your assistance!
Re: Comparative Wealth Index [message #8921 is a reply to message #8910] Thu, 14 January 2016 15:47 Go to previous messageGo to next message
Liz-DHS
Messages: 1290
Registered: February 2013
Senior Member
Dear User,
A response from Dr. Shea Rutstein,
Quote:
Since the DHS Wealth Index has a mean of zero by construction, applying the constant would give a mean of -0.932 for the CWI, not zero! The beta term just expands or reduces the dispersion (equally) so it doesn't change the mean. Has the user reversed the coefficients? Has s/he applied weights (should not for evaluating the mean)? Subset the data?

[Updated on: Mon, 25 January 2016 13:35]

Report message to a moderator

Re: Comparative Wealth Index [message #9064 is a reply to message #8921] Wed, 03 February 2016 21:35 Go to previous messageGo to next message
nael is currently offline  nael
Messages: 6
Registered: January 2016
Member
Dear Liz-DHS,

Thanks a lot for the update! It helped me figure out the mistake I did in the calculations.

Sorry for the inconvenience!
Re: Comparative Wealth Index [message #9149 is a reply to message #9064] Tue, 16 February 2016 12:01 Go to previous messageGo to next message
Liz-DHS
Messages: 1290
Registered: February 2013
Senior Member
Dear User,
Dr. Rutstein reviewed his responses to you and asked that you used the attached file containing the CWI coefficients. What you had seemed fine except for 2010. Should use the attached.
Thanks!

[Updated on: Tue, 16 February 2016 12:04]

Report message to a moderator

Re: Comparative Wealth Index [message #10713 is a reply to message #8868] Mon, 05 September 2016 10:36 Go to previous message
Jan Eberle is currently offline  Jan Eberle
Messages: 2
Registered: September 2015
Member
Hello everyone,

is it possible to obtain an updated list of CWI coefficients also containing more recent surveys?

Thanks!


Previous Topic: Comparative Wealth Index - detailed instructions for computation
Next Topic: Wealth index Construction for Burundi 2010
Goto Forum:
  


Current Time: Sat Aug 18 22:22:10 Eastern Daylight Time 2018