| Re: Children ever death and children ever died [message #30469 is a reply to message #30467] |
Wed, 04 December 2024 14:41  |
Janet-DHS
Messages: 938
Registered: April 2022
|
Senior Member |
|
|
Following is a response from DHS staff member, Tom Pullum:
The under-5 mortality rates (in chapter 8 of most final reports) describe the probability that a birth will result in a death within the first month, year, 5 years, etc. Roughly speaking, the number of child deaths is proportional to the number of births. If there are fewer births, there will be fewer deaths, just because of that proportionality. With just v201 and v206+v207, you can't go beyond that kind of proportionality. And with just those variables you are really limited, because they don't tell you the age at death for children who died.
Perhaps you are trying to get at some effect beyond that proportionality. In chapter 8 you will see a section on so-called high-risk births, mainly closely spaced births. As fertility declines, there tend to be fewer high-risk births, which also reduces the number of deaths. Then there are separate trends in the under-five mortality rates than are separate from the fertility trends. Historically, there have been declines in mortality separately from declines in fertility.
With the retrospective birth histories, you can get at most of these effects. You can also get at whether women/couples tend to replace a child who has died, an effect in the opposite direction that I mentioned earlier.
So, I would say that your research question is good for DHS data if you use the birth histories, but not if you just use v201 and v206+v207. I don't have anything to add.
|
|
|
|
Members
Search
Help
Register
Login
Home
