Following is a response from Senior DHS staff member, Tom Pullum:
Because of the subsampling for biometry in this survey, about a third of women were not asked these questions. They must be excluded from the denominator. The following Stata lines will match the 13.8% in Table 12.9:
use "...BDIR81FL.DTA", clear
tab1 v744*,m
gen outcome=0 if v744a<.
foreach ll in a b c d e {
replace outcome=1 if v744`ll'==1
}
tab outcome [iweight=v005/1000000]
I will alert the staff in charge of STATcompiler that this indicator was omitted. I believe that was just an oversight.
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
