| Comparing national-regional and state [message #27916] |
Sat, 21 October 2023 22:18  |
Lincoln
Messages: 31
Registered: July 2022
|
Member |
|
|
Dear forum,
I am trying to use the HR file for 2019-21 round to compare (mean and proportions) the data between one state, group of states and national data for disability: any one died in the house hold? how many died?
Please help, using STATA.
Looking forward for kind help
Regards
Lincoln
|
|
|
|
| Re: Comparing national-regional and state [message #27934 is a reply to message #27916] |
Mon, 23 October 2023 16:46  |
Bridgette-DHS
Messages: 3043
Registered: February 2013
|
Senior Member |
|
|
Following is a response from Senior DHS Stata Specialist, Tom Pullum:
This is potentially a complex analysis. The following lines will tell you the number of deaths in the household and whether there were any deaths at all:
use "C:\Users\26216\ICF\Analysis - Shared Resources\Data\DHSdata\IAHR7EFL.DTA", clear
* a death is indicated by sh90_1 through sh90_5 being less than .
gen ndeaths=0
forvalues li=1/5 {
replace ndeaths=ndeaths+1 if sh90_`li'<.
}
gen anydeaths=0
replace anydeaths=1 if ndeaths>0
tab ndeaths anydeath [iweight=hv005/1000000]
However, this describes all deaths since January 2016, the beginning of the reference period for the questions. The length of the reference period is much longer for the households that were interviewed later than for those that were interviewed earlier. Are you going to restrict to an interval such as the last 36 months? Are you going to include all combinations of age and sex? Are you going to try to take exposure to risk into account, by using the household age-sex composition? And how are you going to relate deaths to disabilities? The survey does not include any disability information about the household members who died. Have you looked at the tables in the final report? They include the CDR for each state. Let us know if you have other questions.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
