The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Mortality » calculating intersectional mortality rates (problems in editing do file from GITHUB)
calculating intersectional mortality rates [message #22527] Thu, 25 March 2021 08:41 Go to next message
shivani is currently offline  shivani
Messages: 3
Registered: March 2021
Member
Hi
I wish to calculate the mortality rates for children as well as adults using the NFHS datasets. Since it's a complex procedure to match the methodology used by DHS in its reports, I rescued to the do file by GITHUB (also mentioned by DHS representative on the forum). However, the do file gives estimates for only a range of covariates (say age, gender and so).

I, on the other hand, wish to calculate mortality rates for a range of intersections among the covariates. For instance, caste and gender: I wish to calculate the mortality rates for males and females belonging to different caste groups.

Or say, across wealth quintiles among caste groups....



Can you please assist with how to do so?

What changes in the do file are needed??
Re: calculating intersectional mortality rates [message #22528 is a reply to message #22527] Thu, 25 March 2021 09:19 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2270
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

The current versions of the GitHub programs are only designed to match the tables in the main reports. It would be possible to modify the code. However it will be easier just to reduce the data file to the subpopulation you want. For example, if you open the BR data file, then enter "keep if v106==0 & v190==1", save the file (with a modified filename) and produce estimates of the under-five mortality rates for that file, the estimates will refer to children whose mothers have no education and whose households are in the lowest wealth quintile. This will be tedious, but will produce what you want. Note that the sample sizes may be very small for some combinations.
Re: calculating intersectional mortality rates [message #22593 is a reply to message #22528] Thu, 08 April 2021 08:19 Go to previous messageGo to next message
shivani is currently offline  shivani
Messages: 3
Registered: March 2021
Member
hey!
Thanks for the reply. i was able to calculate the desired statistics because of that suggestion.
However, when I use the same do file on NFHS 3 datasets, its giving me an error code of "No observations". i havent changed anything except for the modifications in the names of covariates.

what shall be done? please assist
Re: calculating intersectional mortality rates [message #22622 is a reply to message #22593] Sat, 10 April 2021 15:13 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2270
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

If you are restricting the sample, you have to avoid small subpopulations. However, I would not expect this error message. All I can suggest is that you check that you have not made an error in how you reduced the sample.

Previous Topic: s572 variable
Next Topic: Calculating perinatal mortality rate for each of the five years preceding the survey
Goto Forum:
  


Current Time: Wed May 12 16:50:50 Coordinated Universal Time 2021