Re: Merging Male with Birth data? [message #25333 is a reply to message #25331] |
Wed, 05 October 2022 16:11 ![Go to previous message Go to previous message](/theme/default/images/up.png) |
Bridgette-DHS
Messages: 3150 Registered: February 2013
|
Senior Member |
|
|
Following is a response from DHS staff member, Tom Pullum:
The child records in the BR files already include almost all of the mother's information that is in the IR file. If you think some variables are missing and you still want to merge with the IR file, you match cases in the IR and BR with the same v024 v001 v002 v003 in both files. You keep if _merge==3.
Merging the BR and MR files is not a good idea. The first problem is that there was only a 1/6 subsampling of males from the PR file. 1/2 of the households in 1/3 of the clusters were selected for the men's survey. that means you could only match at most 1/6 of the children with their father. Second, you can only match with the father if both the child and the father were present the same household. You can find the child in the household by matching v024 v001 v002 v003 b16 in the BR file with hv024 hv001 hv002 hvidx in the PR file. The father's line number, if in the household, is hv114. You then merge hv024 hv001 hv002 hv114 with mv024 mv001 mv002 mv003 in the MR file. The code to merge children with fathers has been given elsewhere in the forum.
If you want children under 5, use the KR file rather than the BR file. If you use the BR file, the older children are less likely to be in the household. Both the KR and BR files include children who have died, and you will not find them in the household survey.
|
|
|