| Re: 2022 DHS FGM Prevalence Question [message #27472 is a reply to message #27440] |
Fri, 18 August 2023 08:54  |
Bridgette-DHS
Messages: 3199
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS staff member, Tom Pullum:
You are on the right track, but it would be preferable to pool the counties using the n's in the data.
I have not used that PowerPoint, but have re-calculated the FGM prevalence by county. (Chapter 18 of the final report does not include county estimates). The Stata lines below show how this is done. The text file "FGM_Kenya_2022_DHS.txt" gives a log file with the results. The pooled estimate for those two counties combined is 65.56%. This could be calculated as a weighted average of 51.01% and 77.29% for the two counties, with n's of 373.65 and 463.22. The n's are weighted frequencies (numbers of women in the sample) and that's why they are not integers. These n's refer to the number of women age 15-49, not the total population.
Rounding the percentage, I get 66%, compared with your 65%. Virtually the same.
use "...KEIR8AFL.DTA" , clear
tab1 g100-g103 [iweight=v005/1000000],m
gen fgm=.
replace fgm=0 if g100==0 | g102==0
replace fgm=1 if g102==1 | g103==1
tab fgm [iweight=v005/1000000]
* convert to a percentage
replace fgm=100*fgm
* get the percentages and n's by counties
tab v024 fgm [iweight=v005/1000000], row
* estimates for Narok + Kisii
tab v024 fgm [iweight=v005/1000000] if v024==33 | v024==45, row
* pooled estimate for Narok + Kisii
tab fgm [iweight=v005/1000000] if v024==33 | v024==45
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
