* Program to calculate table 18.6 in the Nigeria 2018 report * Circumcision of girls age 0-14 * This program does not include covariates; it just shows how to calculate the indicator * Apart from having to do a merge with a change of index, there are two main complications: * The FGC questions were only asked for a subsample (about 1/3) and the analysis must be * limited to the subsample with g100=0 or g100=1 (remove if g100=.) * All the daughters age 0-14 must be in the denominator, including those whose mothers have * g100=0; just those with g121=1 go into numerator * the numerator * PREPARE THE IR FILE set more off use v001 v002 v003 gidx_* g121_* using NGIR7AFL.DTA , clear rename *_0* *_* * RESHAPE THE IR FILE SO THERE IS ONE RECORD PER DAUGHTER reshape long gidx_ g121_, i(v001 v002 v003) j(sequence) drop sequence rename *_ * drop if gidx==. rename gidx bidx gen in_IR=1 sort v001 v002 v003 bidx save NGIRtemp.dta, replace * PREPARE THE BR FILE use NGBR7AFL.DTA , clear * Identify girls, living and age 0-14 (b15==1 is redundant) keep if b4==2 & b5==1 & b8<=14 * Crucial line to drop the mothers and daughters who did not get the long questionnaire * drop the girl if the g question were not asked of her mother drop if g100==. keep v001 v002 v003 v005 bidx b8 gen age5=1+int(b8/5) label define age5 1 "0-4" 2 "5-9" 3 "10-14" label values age5 age5 * in_BR identifies a daughter who is eligible for a g121 code gen in_BR=1 * MERGE THE BR FILE WITH THE RESHAPED IR FILE sort v001 v002 v003 bidx merge v001 v002 v003 bidx using NGIRtemp.dta tab _merge drop _merge label define g121 0 "No" 1 "Yes" label values g121 g121 * Some girls in the BR file do not have a value on g121 because their mothers had not heard of fgc. * Crucial line to get the correct denominator replace g121=0 if in_IR==. & in_BR==1 * Unweighted tab age5 g121,m * Weighted tab age5 [iweight=v005], summarize(g121)