Home » Countries » Other countries » Côte d'Ivoire 2021 (Missing values hw72)
| Côte d'Ivoire 2021 [message #30048] |
Mon, 16 September 2024 12:44  |
lillo?S
Messages: 26
Registered: December 2015
|
Member |
|
|
|
I am using the latest DHS dataset for Côte d'Ivoire and am specifically interested in the data on wasting (i.e., variable hw72) from the KR file (CIKR81FL.DTA). I have noticed that the wasting information is missing for 5,767 children and ".a" for 57 children. Could you please provide some insights into the missingness patterns for this variable? Thank you very much!
|
|
|
|
|
|
|
|
| Re: Côte d'Ivoire 2021 [message #30144 is a reply to message #30087] |
Thu, 03 October 2024 14:14  |
Janet-DHS
Messages: 888
Registered: April 2022
|
Senior Member |
|
|
Following is a response from DHS staff member, Tom Pullum:
It is not actually a problem that many of the children in the KR file are missing a value of hw1 (months of age). This survey had subsampling for anthropometry. hw1 is only available in the subsample for anthropometry, and is missing ( a dot for NA) otherwise. Use b19, which is available for all children in the KR file, for age in months.
In CIPR81FL.dta I see 10,595 cases of children under 5 (with HV105 = 0-4). I see 5,620 cases with HC1 = 0-59 months. HC1 is an anthropometry variable and is only available for the subsample that was used for anthropometry. I assume that was half of the households, but I would need to check the report.
In CIKR81FL.dta I see 10,645 unweighted cases. I see 4,902 unweighted cases with HW1 = 0-59 months. Again, this is because anthropometry was only in a subset of cases (half of households I'm guessing). For B19 I'm also seeing 10,645 cases.
If I use the following code:
clonevar hv001 = v001
clonevar hv002 = v002
clonevar hvidx = b16
merge m:1 hv001 hv002 hvidx using "C:\Users\21180\OneDrive - ICF\Data\DHS_Stata\CIPR81FL.dta"
keep if _merge==3 | _merge==1 | inrange(hv105,0,4)
tab _merge
I get the following:
. tab _merge
Matching result from |
merge | Freq. Percent Cum.
------------------------+-----------------------------------
Master only (1) | 1,276 10.62 10.62
Using only (2) | 1,373 11.42 22.04
Matched (3) | 9,369 77.96 100.00
------------------------+-----------------------------------
Total | 12,018 100.00
If I tab by b5, I see 697 children of respondents who have died, plus 579 who do not match with the PR file as they do not live in the interviewed household. There are also 1373 children age 0-4 who are not children of respondents and thus are not listed in the respondent's pregnancy history.
. tab b5 _merge,m
child is | Matching result from merge
alive | Master on Using onl Matched ( | Total
-----------+---------------------------------+----------
no | 697 0 0 | 697
yes | 579 0 9,369 | 9,948
. | 0 1,373 0 | 1,373
-----------+---------------------------------+----------
Total | 1,276 1,373 9,369 | 12,018
If I tab by b16, I see
. tab b16 _merge,m
child's line number | Matching result from merge
in household | Master on Using onl Matched ( | Total
----------------------+---------------------------------+--- -------
not listed in househo | 579 0 0 | 579
2 | 0 0 204 | 204
3 | 0 0 1,878 | 1,878
4 | 0 0 1,831 | 1,831
5 | 0 0 1,583 | 1,583
6 | 0 0 1,170 | 1,170
7 | 0 0 849 | 849
8 | 0 0 561 | 561
9 | 0 0 382 | 382
10 | 0 0 295 | 295
11 | 0 0 196 | 196
12 | 0 0 142 | 142
13 | 0 0 85 | 85
14 | 0 0 67 | 67
15 | 0 0 42 | 42
16 | 0 0 26 | 26
17 | 0 0 17 | 17
18 | 0 0 12 | 12
19 | 0 0 7 | 7
20 | 0 0 3 | 3
21 | 0 0 7 | 7
22 | 0 0 3 | 3
23 | 0 0 3 | 3
24 | 0 0 1 | 1
25 | 0 0 1 | 1
26 | 0 0 2 | 2
27 | 0 0 1 | 1
33 | 0 0 1 | 1
. | 697 1,373 0 | 2,070
----------------------+---------------------------------+--- -------
Total | 1,276 1,373 9,369 | 12,018
The match looks as expected. Let us know if you still have questions.
|
|
|
|
Goto Forum:
Current Time: Fri Nov 22 23:45:22 Coordinated Universal Time 2024
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
