The DHS Program User Forum      
Discussions regarding The DHS Program data and results
Home » Topics » General » Living Arrangement in the PR file
Living Arrangement in the PR file [message #10782] Mon, 12 September 2016 06:16 Go to next message
waltine is currently offline  waltine
Messages: 20
Registered: March 2013
Location: South Africa
Member
Dear All,

I am trying to examine the relationship between living arrangement measured by children living with both parents, living with mother alone, living with father alone and living with neither of the two. I am using the PR recode because it also has nutritional status which is another variable of interest. I have been able to successfully merge the PR and the KR recode. But after merging, I find that the variable created (living arrangement) does not include two categories (children living with father alone and children with no parents). I am assuming this is because the KR is asked from mothers alone. Could this be the case? If yes, what can I do or can I use another recode.

I am currently working with the Nigerian 2013 dataset.


waltine
Re: Living Arrangement in the PR file [message #10799 is a reply to message #10782] Thu, 15 September 2016 11:58 Go to previous messageGo to next message
Trevor-DHS is currently offline  Trevor-DHS
Messages: 535
Registered: January 2013
Senior Member
Yes, your assumption is correct. If you start from the PR file and merge in the KR file and only use those that match, you will only have cases where the mother was alive and interviewed. In our tabulations, and specifically the table on children's nutritional status, we use all de facto children (who stayed in the household the previous night), irrespective of whether the mother was interviewed or not. For the other characteristics, we use the data from the mother via the KR or IR file if the mother was interviewed, and from the PR file (match on the mother's line number) if the mother was in the household, but not interviewed, and we set the characteristics to missing for mothers who do not live in the household or are dead. Thus we keep all of de facto children from the PR file, irrespective of whether they match with the KR file.
Re: Living Arrangement in the PR file [message #10803 is a reply to message #10799] Sat, 17 September 2016 12:40 Go to previous messageGo to next message
waltine is currently offline  waltine
Messages: 20
Registered: March 2013
Location: South Africa
Member
Hello Trevor,

Thanks for the response. Does this mean I can still use the PR file based on this recommendation "we set the characteristics to missing for mothers who do not live in the household or are dead. Thus we keep all of de facto children from the PR file, irrespective of whether they match with the KR file. ".... How do I set the characteristics to missing? Would this be before or after merging?


waltine
Re: Living Arrangement in the PR file [message #10804 is a reply to message #10803] Sat, 17 September 2016 18:05 Go to previous messageGo to next message
Trevor-DHS is currently offline  Trevor-DHS
Messages: 535
Registered: January 2013
Senior Member
The steps are to 1) start wit the PR file, 2) merge the KR file, keeping all of the PR file cases (i.e. not dropping the PR cases that do not match), 3) for any case from the PR that did not merge with the KR file, check if the mother is alive and lives in the household (see hv112), and if so merge in the characteristics of the mother, 4) for any case that still does not match, the value of the characteristics will generally be missing. For step 3) you will have to create a small file from the PR file of just the characteristics you need from the mother (e.g. we take the women's education) and in that file you will have to match hv112 (mother's line number) from the original PR file with hv101 (person's line number - identifying the mother) in the smaller extract from the PR file.
Re: Living Arrangement in the PR file [message #10849 is a reply to message #10804] Sat, 24 September 2016 09:43 Go to previous message
waltine is currently offline  waltine
Messages: 20
Registered: March 2013
Location: South Africa
Member
Thanks Trevor.

I am getting stuck at step 2. I don't understand how to go about step 3.


waltine
Previous Topic: Percentage of women who gave birth by age 20
Next Topic: Ethics documents for Zambia
Goto Forum:
  


Current Time: Fri Nov 17 21:40:48 Eastern Standard Time 2017