Marriage before 18 years [message #30435] |
Thu, 28 November 2024 06:26 |
marian
Messages: 14 Registered: August 2019
|
Member |
|
|
Hi,
Please help with the code for marriage before 18 years among women aged 20-24 years, as shown in the NFHS 5 reports. I have used the below code, but the data is not matching factsheet reports:
gen national_wgt = v005 / 1000000
recode v511 (.=0) (0/17 = 1 "yes") (18/49 = 0 "no"), gen (ms_afm_18)
replace ms_afm_18 = . if v012<18
label var ms_afm_18 "First marriage by age 18"
tab v024 ms_afm_18 [iw=national_wgt] if v012>19 & v012<25, row
Please help.
|
|
|
Re: Marriage before 18 years [message #30445 is a reply to message #30435] |
Mon, 02 December 2024 06:21 |
Bridgette-DHS
Messages: 3203 Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
Your Stata code is correct. For a test, I applied it to the Philippines 2022 data, and matched the report (Table 4.3) with 9.4% (for the total country). When I apply it to the NFHS-5 data I get 25.3% (is this what you were getting?), but the final report, Table 6.2, gives 23.3% and STATcompiler gives 22.3%. I cannot even match the frequencies ("Number of respondents") given in Table 6.2 of the final report.
In general, STATcompiler estimates are preferable to final report estimates, because they are updated when the data files are updated, and IAIR74FL.dta is the 5th version (0, 1, 2, 3, 4) of the IR file for this survey. I am surprised that I cannot match an estimate on STATcompiler, but that's how it is. It's up to you to decide what to do--but as I said, your Stata code is correct.
|
|
|