|
Re: TFR and ASFR at state level using TFR2 in STATA [message #8908 is a reply to message #8906] |
Tue, 12 January 2016 20:43   |
Reduced-For(u)m
Messages: 292 Registered: March 2013
|
Senior Member |
|
|
I believe that if you want to use TFR2 to calculate fertility rates at the regional level, you could just open your dataset, keep the observations from one region, run TFR2, get the results, and then repeat for other regions. I do not believe (at least since I last checked) that TFR has a specific command to use sub-samples, but you don't really need it - just keep the observations from one region, run and repeat for the other regions.
You should definitely check the sample sizes for each region though, because these estimates are likely to be fairly noisy (and check the DHS report for the Pakistan dataset you are using to make sure that it was designed to be representative at the regional level).
|
|
|
|
Re: TFR and ASFR at state level using TFR2 in STATA [message #9337 is a reply to message #8915] |
Tue, 15 March 2016 07:53  |
owraza
Messages: 31 Registered: December 2013 Location: Tehran
|
Member |
|
|
In doing sub-group analysis, you need to take care of 'all women factor'. The method described above is completely valid as far as you manage to use correct variable for all women factor in that split file.
Other way of doing sub-group analysis, as described by author of tfr2, is to use following code line:
use "C:\Users\PKIR61FL.DTA", clear
by v024, sort: tfr2 [pweight=v005], len(3) ageg(5) bvar(b3_*) dates(v008) wbirth(v011) awf( awfactr )
Regards
|
|
|