Home » Topics » Reproductive Health » Benin 2011-2012 Daughter's circumcision status (Respondent-reported daughter's circumcision status)
Benin 2011-2012 Daughter's circumcision status [message #25179] |
Fri, 09 September 2022 13:41 ![Go to next message Go to next message](/theme/default/images/down.png) |
Lindsey
Messages: 2 Registered: September 2022
|
Member |
|
|
I am using data from the Benin 2011-2012 survey on the respondent-reports of daughters' female genital cutting (FGC) status. The survey asks respondents to report FGC status for daughters up to age 15, but there is information for daughters up to age 32.
The following are caseids with FGC status reported for daughters that are over 15:
"8 5 2"
"28 14 7"
"28 20 3" - gidx (the variable that links the daughter to the respondent's birth history variable, bidx) has 3, 5, 11, 12, but the respondent's children in bidx 1, 4, 8, and 9 are also female. So did the respondent not report FGC status for those daughters? She reports the FGC status for 4 daughters, and her fourth youngest daughter does meet the cutoff for FGC reporting: she's 14. But child in bidx place 12 (for whom gidx is reported) is 26 years old.
"34 17 2" - gidx has 3, 6, 7, but kid 5 is also female. Kid 3 is 18 years old and Kid 7 is 23.
"31 13 2" - gidx has 1, 10, 11, but kids 2 and 5 are also female. Only girls 1 and 2 meet the age requirement. Kid in bidx 5 is 19 years old and kid 11 is 31.
There are other caseids, but I think those give a good idea of the issue.
Is it the case that some enumerators ignored the age restriction (which would not a problem for my project), or do you think there could be an issue with the gidx/bidx matching (more concerning for my project)? There are multiple enumerators with this issue in Benin 2011-2012, but no other datasets have this issue. That is, other datasets are consistent with the age restriction stated in the survey.
Thanks for your help!
[Updated on: Fri, 09 September 2022 13:42] Report message to a moderator
|
|
|
Re: Benin 2011-2012 Daughter's circumcision status [message #25194 is a reply to message #25179] |
Mon, 12 September 2022 17:05 ![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) |
Janet-DHS
Messages: 816 Registered: April 2022
|
Senior Member |
|
|
Following is a response from DHS staff member Tom Pullum:
I looked in detail at the last case you describe: "31 13 2 - gidx has 1, 10, 11, but kids 2 and 5 are also female. Only girls 1 and 2 meet the age requirement. Kid in bidx 5 is 19 years old and kid 11 is 31."
This is the case in BJIR61FL.dta with v001==31, v002==13, v003==2. I reduced the file to just this case and the b*_* and g*_* variables and then looked at how these two sets of variables match up.
Yes, gidx_01 is 1, gidx_02 is 10, and gidx_03 is 11, identifying cases 1, 10, and 11 in the birth histories. If you then enter these three lines:
list b4_*, table clean
list b5_*, table clean
list b8_*, table clean
you will get a readable list of the sex (b4), survival status (b5), and age (b8) of all the children this woman had. You will see that kids 2 and 5 are female, yes, BUT they have b5=0, that is, they were not survivors. To be eligible for the FGC questions the daughter must be living. So no problem there. You just didn't check the survival status.
Kids 10 and 11 are daughters and are surviving, but their ages are 26 (not 19) and 31. They were not eligible for the FGC questions. This does appear to me to be an error during data collection.
I think the other cases you cite are similar to this one, in that the questions may have been asked about daughters who were out of range in terms of current age. Looking at item 1109 in the questionnaire (in an appendix to the final report) I see that the questions were to be asked only about living daughters born in 1996 or later.
Although I believe that you have indeed identified examples of a type of error that occurred during fieldwork, I am sure that superfluous cases such as these were not included in the tables in the report. In general, the construction of an indicator or a table includes checks that have the effect of removing such cases. The tables on FGC would omit these extra cases, because the daughters are outside the age range 0-14.
So this is good detective work on your part, but such errors during fieldwork would not have carried over to the report.
|
|
|
Re: Benin 2011-2012 Daughter's circumcision status [message #25197 is a reply to message #25194] |
Mon, 12 September 2022 20:20 ![Go to previous message Go to previous message](/theme/default/images/up.png) |
Lindsey
Messages: 2 Registered: September 2022
|
Member |
|
|
Thank you very much for the thorough response. This is very helpful. You're right, I didn't think to look at the survival status. Thanks for pointing that out. And it is helpful to know that some enumerators may have asked about daughters above the age description and not that the gidx and bidx variables were misaligned. This is not a problem for me as I can just drop these daughters from my analysis.
|
|
|
Goto Forum:
Current Time: Sat Sep 7 06:18:54 Coordinated Universal Time 2024
|