The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Wealth Index » Urban Score vs. Combined Score
Urban Score vs. Combined Score [message #10862] Mon, 26 September 2016 12:21 Go to next message
ajhollenbach is currently offline  ajhollenbach
Messages: 5
Registered: September 2016
Member
Greetings,

I am currently looking to use the Wealth Index for classifying survey respondents in Indonesia by wealth quintile. I have two questions in this regard:

First, the project that I am supporting will only be conducting surveys in urban areas, and are therefore interested in classifying people by wealth across the urban spectrum only. I see that the excel file for Indonesia's last survey in 2012 included coefficients for both urban and rural households. If we use the coefficients for only the urban households to score each new survey respondent in combination with the existing breakpoints, will this be akin to ranking households along the wealth spectrum of urban areas only? I noted that the spreadsheet also includes a formula for calculating a "combined score" once the urban score is reached(Combined Score= 0.489 + 0.805 * Urban Score), but assume that we would NOT want to use this formula, as we are only interested in urban areas. Appreciate if you could confirm this assumption.

Second, if one was interested to further compare the wealth status of households only with other households in, say, the provinces of Java (which tends to be a bit wealthier overall), is it possible to establish new breakpoints for Java by scoring households from Java only, ranking the scores from highest to lowest, and then dividing the Java-only dataset into quintiles?

Thanks Much.
Re: Urban Score vs. Combined Score [message #10879 is a reply to message #10862] Tue, 27 September 2016 22:34 Go to previous messageGo to next message
Liz-DHS
Messages: 1516
Registered: February 2013
Senior Member
A response from Dr. Shea Rutstein:
Quote:

You are correct on both points. You can use just the urban coefficients for all urban areas, and for Java, use the second approach of estimating Java-specific quintiles. You should point out that these quintiles are not comparable to either the national nor the urban quintiles.

Thank you for your post!
Re: Urban Score vs. Combined Score [message #10884 is a reply to message #10879] Wed, 28 September 2016 08:03 Go to previous messageGo to next message
ajhollenbach is currently offline  ajhollenbach
Messages: 5
Registered: September 2016
Member
Thanks very much for your response. Just to clarify my first question on our focus on urban areas, following the application of the urban coefficients, can you confirm that we should NOT then use the formula to calculate the "combined score" (i.e. Combined Score = 0.489 + 0.805 * Urban Score). I guess that I am just a bit unclear on the rationale behind the "combined score" formula.

Thanks,
Allen Hollenbach
Re: Urban Score vs. Combined Score [message #10887 is a reply to message #10884] Wed, 28 September 2016 11:15 Go to previous message
Liz-DHS
Messages: 1516
Registered: February 2013
Senior Member
Dr. Rutstein's response:
Quote:

The combined score is for the overall national level wealth index and so is not appropriate for just the urban areas. Thus don't use the regressions to combine the urban and rural scores.


Previous Topic: Cooking Fuel - Haiti DHS 2012 -Wealth Index
Next Topic: Brazilian asset index
Goto Forum:
  


Current Time: Tue Mar 19 07:50:04 Coordinated Universal Time 2024