Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: about the temporal desagragation


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: about the temporal desagragation
Date   Wed, 10 Feb 2010 22:20:07 -0000

I would be equally severe on this idea if the proposal were as Maarten imagines. But the Denton method is not quite the kind of statistical fiction-writing implied. Another variable is used in the interpolation, i.e. the method uses, and indeed presupposes, other information on the time scale to be interpolated. 

I can't answer the original question specifically but a wild guess is that inferring quarterly from half-yearly follows similar principles, except that you might need do the details yourself. 

Nick 
n.j.cox@durham.ac.uk 

Maarten buis

The real question is why would you want to do that? The whole
point of empirical science is that we give an answer to a 
question based on what we have seen. If you have seen half
yearly data, then that is the information you'll have to work
with. Interpolation is not going to add information, it is 
just adding "assumptional noise" that will only make the 
link "question --> stuff we have seen --> conclusion" less
clear. 

One possible answer is that you want to use that another 
time series as an explained or explanatory variable which
measured on a quartly basis. In that case you might (multiply)
impute the half yearly data. The aim of such an excercise is 
not to add information, the quartly information on the half 
yearly data is lost forever and no amount of statistical 
trickery can recover it. Instead, the aim is to preserve
the information on the variable that was measured on a 
quarterly basis, which would otherwise have been thrown away 
if you analysed the two timeseries together. If that is your 
aim you could look at -help mi- and -findit ice-. Notice,
that you are still going to add quite a bit of assumptional
noise, so the trade off you will have to make is if the
additional information that you are going to preserve is 
worth the additional assumptional noise.


--- On Tue, 9/2/10, amatoallah ouchen wrote:
> I have a half yearly data that I would like to transform to
> quartly data, is there any command or module for getting
> that result? I've seen that Pr Kit Baum had made a stata
> module to interpolate a quarterly flow series from annual
> totals via proportional Denton method.

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany

http://www.maartenbuis.nl
--------------------------


      

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index