Recalculating and comparing Wealth Index for Ghana [message #1357] |
Mon, 17 February 2014 10:51 ![Go to next message Go to next message](/theme/default/images/down.png) |
Kwesi.Nkrumah
Messages: 4 Registered: March 2013 Location: Ghana
|
Member |
|
|
I Have two questions.
1. I want to recalculate the wealth index for the Ghana Demographic and Health Survey datasest for Ghana for the 1993 and 1998 surveys. I have however looked at the final report for the relevant years and cannot find the variables used. Is it possible to be given the actual variables used. Given that I will be comparing with the 2003 and 2008 surveys, I will equally appreciate it if the variables used for 2003 and 2008 are also added.
Secondly, I realised from the DHS data repository at Measure DHS that the wealth index for 1993 and 1998 have bee calculated and stored in seperate files. Given the need to do some comparisons I will want to use that in addition to my own calculations for robustness checks. My challenge however is how to merge the seperate files containing the wealth index for 1993 and 1998 to their corresponding women's file which am using for the analysis. I have already looked at the section on the Measure DHS website which gives guidance on unique identifiers appropritae for merging different DHS data files. Given that the variables used to compute the wealth index are at the household level, I assume that the seperate wealth index file will be treated as if they were a household file. In that case the unique identifier for merging will be v001 and v002 (i.e. the cluster id and household number). However, I realised that the welth index file has only WHHID which corresponds to the caseid in the women's file and cannot be used to match observations. I will appreciate if I can have some help in merging the seperate wealth index files provided with the women's file.
Thank you very much
Regards
|
|
|
Re: Recalculating and comparing Wealth Index for Ghana [message #1360 is a reply to message #1357] |
Mon, 17 February 2014 13:12 ![Go to previous message Go to previous message](/theme/default/images/up.png) ![Go to next message Go to next message](/theme/default/images/down.png) |
Liz-DHS
Messages: 1516 Registered: February 2013
|
Senior Member |
|
|
Dear User,
Here is a response from one of our experts, Dr. Shea Rutstein:
1. The wealth index variables and their coefficients are attached.
2. The link between the wealth index scores files and the data files is the household identification. Each respondent in a household has the same wealth index score and quintile. To link between the household data file and the wealth index score file, you can use the HHID variable in the HH data file and the WHHID in the wealth index score file. These variables need to be in the same format (usually alphanumeric of 12 characters in length). Generally, the HHID is made up of the cluster number (HV001) and the household number (HV002). The individual identification CASEID has the household member line number of the respondent added to the household id and is generally alphanumeric of 15 characters in length. To match with individual data, you could trim the last 3 characters from CASEID or reconstruct HHID from V001 and V002. In a few surveys, an additional variable will need to be used but not in Ghana 1993 and 1998. Inspect the created variable to see if it matches the WHHID variable. You may have to add/trim spaces at the beginning or the end and the created id variable should have the same length as WHHID.
Thank you!
|
|
|
|