Use of wealth score for regression [message #3645] |
Thu, 22 January 2015 06:33 |
plnep
Messages: 11 Registered: November 2014
|
Member |
|
|
Hi forum users,
Can we use the wealth score available as continuous variable?
gen asset=v191/100000
Just wanted to check if I need to be aware when using it in pooled survey from different years !! I am using Nepal DHS from 2001, 2006, 2011.
Any insights would be greatly appreciated.
Thanks,
Pnep
|
|
|
Re: Use of wealth score for regression [message #3662 is a reply to message #3645] |
Fri, 23 January 2015 12:20 |
Liz-DHS
Messages: 1516 Registered: February 2013
|
Senior Member |
|
|
Dear User, Here is a response from one of our experts, Dr. Shea Rutstein:
Quote:The DHS wealth index is calculated separately for each survey (with a couple of exceptions, not being Nepal) so that the value of the wealth score in one survey may not (usually the case) correspond to the same level of wealth in another survey even in the same country. For that, the Comparative Wealth Index was created. To create the Comparative Wealth Index for those surveys use the following:
CWI score=Alpha + Beta*v191
Country Year Constant (Alpha) Coefficient (Beta)
Nepal, 2001 -1.0876444 0.6696379
Nepal, 2006 -0.8041832 0.8550128
Nepal, 2011 -0.3603931 1.0388343
To create comparable quintiles, use these values as cutpoints:
Cut points for CWI quintiles:
1-2 -0.90802
2-3 -0.38583
3-4 -0.01189
4-5 0.74156
And yes, you can use the values of the CWI (and individual wealth index scores) instead of quintiles.
Shea Rutstein
|
|
|
|
|
|
|
|