caseid in KR file for Nigeria 2018 [message #20171] |
Mon, 05 October 2020 10:11 |
fabpra
Messages: 1 Registered: October 2020
|
Member |
|
|
Dear DHS Program team,
I am struggling with the 2018 Nigeria KR file. I thought I had understood from the guide to DHS statistics that in the KR file, the child is the unit of observation and therefore a caseid is assigned to each single child. Now I realized that the caseid seems to be on the children's mothers.
I checked by sorting the Stata file by v001 v002 v003 - children of the same mother have the same caseid in this dataset.
Where is the error in my reasoning? Thank you in advance for your help.
Best,
Fabienne
|
|
|
|
|
Re: caseid in KR file for Nigeria 2018 [message #26266 is a reply to message #26259] |
Wed, 01 March 2023 11:05 |
Bridgette-DHS
Messages: 3203 Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
A household will often contain more than one child under 5. They could be children of the same mother, in which case v001 v002 and v003 will be the same, or they could be children of different mothers, in which case v001 v002 will be the same but they will have different values of v003 (the mother's line number in the household). It would be possible for two children in the same household to have the same values of v001 v002 and bidx (for example, bidx=1), but if that happens they will have different values of v003. The full unique identifier for children in the KR file is v001 v002 v003 bidx.
You have to decide whether mothers, or children are the units of analysis. I think you may be unsure about that. That's up to you to decide.
|
|
|