The DHS Program User Forum      
Discussions regarding The DHS Program data and results
Home » Data » Merging data files » MERGING INDIVIDUAL RECORD WITH MEN'S RECORD
MERGING INDIVIDUAL RECORD WITH MEN'S RECORD [message #13142] Wed, 27 September 2017 17:07 Go to next message
Christopher is currently offline  Christopher
Messages: 4
Registered: February 2017
Location: taipei, Taiwan
Member
Dear Seniors,

Am trying to find the impact of men involvement in the uptake of Antenatal care for the Malawi DHS trying to compare the three data sets of 2004,2010 &2016. hence i want to merge the Individual records for the year and the Men's record; i tried to use the couple records but has only those records of the married individuals but i also want to have a general overview of all women in the sample.

Help me out with the SAS command that i may use in merging the Individual Record (mwir7hsd) & Mens Record (mwmr7hsd) for Malawi dataset.

Thank you.
Re: MERGING INDIVIDUAL RECORD WITH MEN'S RECORD [message #13157 is a reply to message #13142] Fri, 29 September 2017 14:18 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1161
Registered: February 2013
Senior Member

Following is a response from Senior DHS Stata Specialist, Tom Pullum:


I use Stata, not SAS, but will describe the logic of this. We are talking about the IR (women's) file and the MR (men's file). I don't know quite what you mean by "merge". Are you thinking of the household as the unit of analysis, and you want to combine all the men and all the women who live in that household into a single record? For women, the household is identified by v001 v002, and for men by mv001 mv002. Something could be done but there can be several women and several men in the same household and it would be difficult to combine all those records.

If you want a single file that includes both women and men, but they can be on separate records, then "append" is the command we would use in Stata. You would, say, change the mv prefix in the men's data to a v prefix, add a code sex=1 to the men's file, add sex=2 to the women's file (using the same codes as hv104), and then add one file to the other. You have to be careful with the weights, especially if the men were sub-sampled.

If you want coresident women and men partners/spouses on a single record, then (as you note) you just use the couples file. I don't know what other options there may be. Perhaps other users can suggest something else.

Previous Topic: Couple file from male and female dataset
Next Topic: urban/rural variable with education data in dhs in stata
Goto Forum:
  


Current Time: Mon Dec 11 06:15:14 Eastern Standard Time 2017