01 Add control vars and new vars per instructions

[robin-wang]

TO DO FOR DHS

• [x] Socioeconomic inequity control variables in healthcare access and health outcomes in planned analytical work

• Water source (hv201 in DHS HH dataset, already coded for MICS)

• Toilet type (hv205 “”, already coded for MICS)

• Birth order (bord in DHS and brthord in MICS birth dataset)

• Marrital Status - 1 if woman and mother currently married or living in union, 0 otherwise (v501 in DHS and ma1 in MICS woman dataset) – i.e. have it for both woman and child level observations

• Religion of woman and mother (v130 in DHS woman and hc1a MICS household dataset) – i.e. have it for both woman and child level observations

• [x] age of the mother at the child’s birth? (at the child level observation)

• [x] also include the line number of the mother for each kid, that will make future additions of maternal information easier.

• [x] add the following indicators to all standardized micro-datasets (should be quick as they only use variables already produced by the dofiles, so just have to copy-paste the below lines in the right spot in each recode’s/wave’s dofile):

Child under five both stunted and underweight, i.e.: gen c_stuund = (c_stunted == 1 & c_underweight ==1) ; replace c_stunted == . if c_stunted == . | c_underweight == . ; label var c_sttund “1 if child both stunted and underweight” Child did not receive any vaccinations: gen c_vaczero = (c_measles == 0 & c_polio1 == 0 & c_polio2 == 0 & c_polio3 == 0 & c_bcg == 0 & c_dtp1 == 0 & c_dtp2 == 0 & c_dtp3 == 0) ; forvar c_measles c_polio1 c_polio2 c_polio3 c_bcg c_dtp1 c_dtp2 c_dtp3: replace c_vaczero = . if X == . ; label var c_vaczero “1 if child did not receive any vaccinations”

TO DO FOR MICS6

• [x] Socioeconomic inequity control variables in healthcare access and health outcomes in planned analytical work

• Water source (hv201 in DHS HH dataset, already coded for MICS) 20 hh sanitation

• Toilet type (hv205 “”, already coded for MICS) 20 hh sanitation

• Birth order (bord in DHS and brthord in MICS birth dataset) RW: Added to 10 child mortality

• Marrital Status - 1 if woman and mother currently married or living in union, 0 otherwise (v501 in DHS and ma1 in MICS woman dataset) – i.e. have it for both woman and child level observations RW: used ma1 rather than mstatus in wm.dta. This is implemented in 4 sexual health.

• Religion of woman and mother (v130 in DHS woman and hc1a MICS household dataset) – i.e. have it for both woman and child level observations !!! Could we rightly assume head of household represent women/children well?

Ke: religion might be hard since there is different charateristics in the religion and different countries are using different code - religion system

• [x] age of the mother at the child’s birth? (at the child level observation)

• Ke: coded as variable mat_ageatbrth in file 19_child_maternal_edu.do

• RW corrected to c_magebrt, for downstream processing, but used Ke's suggestion for insertion, moved to 10 child mortality

• equals to variable magebrt in bh file

• [x] also include the line number of the mother for each kid, that will make future additions of maternal information easier.

• c_motherln in 9 child anthro

• [x] add the following indicators to all standardized micro-datasets (should be quick as they only use variables already produced by the dofiles, so just have to copy-paste the below lines in the right spot in each recode’s/wave’s dofile):

Child under five both stunted and underweight, i.e.: gen c_stuund = (c_stunted == 1 & c_underweight ==1) ; replace c_stunted == . if c_stunted == . | c_underweight == . ; label var c_sttund “1 if child both stunted and underweight” Child did not receive any vaccinations: gen c_vaczero = (c_measles == 0 & c_polio1 == 0 & c_polio2 == 0 & c_polio3 == 0 & c_bcg == 0 & c_dtp1 == 0 & c_dtp2 == 0 & c_dtp3 == 0) ; forvar c_measles c_polio1 c_polio2 c_polio3 c_bcg c_dtp1 c_dtp2 c_dtp3: replace c_vaczero = . if X == . ; label var c_vaczero “1 if child did not receive any vaccinations”

【RW: Corrected Sven's code in programs. DPT. spelling of stuund.】

[Ortsang]

@robin-wang might need to take a look on the religion part. different country is using very differrent country code, some of the country only contains christianity religions .ie. Tonga and some includes also islamic religions.

[robin-wang]

Status: RW looking into religion.

[robin-wang]

We temporarily use hc1a from hh dataset. This is an indicator for head of household, might be less accurate than the [religion] variable. [Religion] is not always available unfortuantely. There is a leap of faith.

[robin-wang]

Variables added for MICS6. Pending test run and code upload.

[robin-wang]

DHS implemented in DHS VII template on 1104. Sven posted response on the existing MICS6 implementation of edits, to-dos are therefore listed in a new issue. Closed.