Re: Which wealth index to use for state specific studies [message #26026 is a reply to message #25995] |
Mon, 30 January 2023 09:05 |
Janet-DHS
Messages: 893 Registered: April 2022
|
Senior Member |
|
|
Following is a response from DHS staff member, Tom Pullum:
The usual wealth index (hv270, v190, and mv190) gives quintiles at the national level. The urban and rural versions give quintiles within the urban and rural subpopulations, respectively. The state version give quintiles within the states.
The difference between the different versions is just in where the four cut points (to give five quintiles) are placed on the continuous scale (hv271 / v191 / mv191). The components and the weights (or loadings) for the different components are the same. DHS does not construct the continuous scale (the first principal component of a principal components analysis) separately within all these sub-populations.
I recommend the national version for a national-level analysis or if you want to compare states. However, if you are doing a within-state analysis, focusing on some outcome within a specific state, use the state version. I think you just need to ask whether, for example, it is more useful to know whether a household is in the top or bottom wealth quintile (a) nationally, or (b) in the urban or rural subpopulation, or (c) within the state.
|
|
|