The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Weighting data » Re: Applying sample weights to a subset of DHS data
Re: Applying sample weights to a subset of DHS data [message #9399 is a reply to message #9362] Thu, 24 March 2016 09:03 Go to next message
bensonjoon@gmail.com is currently offline  bensonjoon@gmail.com
Messages: 1
Registered: March 2016
Location: Cape town
Member
I want to create the variable fam_struct to assess the adolescent family structure using household member recode data set. This variable has four modalities: 1. living with both parents, 2. live with mother only, 3. live with father only, 4. Not living with both parents. The variable will be merged with individual dataset. May you please help me how I should go about this.
Re: Applying sample weights to a subset of DHS data [message #9529 is a reply to message #9399] Fri, 08 April 2016 14:05 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3017
Registered: February 2013
Senior Member
Following is a response from Senior DHS Stata Specialist, Tom Pullum:

If I understand your question, you can do this very easily, without a merge, because of the questions about parental survival and coresidence that are asked about all children age 0-17 in almost all DHS surveys since about 2000. I use the following lines to describe coresidence with parents. I call the variable "cores_type" rather than "fam_struct" but my categories are the same as yours.

* coresidence typology
tab1 hv112 hv114
gen cores_type=.

* hv112r and hv114r identify parents who are in the household

gen hv112r=0
replace hv112r=1 if hv112>0 & hv112<98

gen hv114r=0
replace hv114r=1 if hv114>0 & hv114<98

replace cores_type=1 if hv112r==1 & hv114r==1
replace cores_type=2 if hv112r==1 & hv114r==0
replace cores_type=3 if hv112r==0 & hv114r==1
replace cores_type=4 if hv112r==0 & hv114r==0

#delimit ;
label define cores_type 1 "Living with both parents" 2 "With mother, not father"
3 "With father, not mother" 4 "Living with neither parent";
#delimit cr

label values cores_type cores_type

To match the DHS table on orphanhood and coresidence, you also must check that the mother and/or father are de jure residents of the household, i.e. that hv102 is 1 for the mother and/or the father. For your purposes, that is probably not necessary.
Previous Topic: weighting data for a subset of main data file
Next Topic: Do I need to weight data for cohort analysis?
Goto Forum:
  


Current Time: Thu Mar 28 20:55:36 Coordinated Universal Time 2024