On 22 Samh 2005, at 10:16, E. Padoano wrote:
. As we are not statistics experts, we received advice from
a colleague that told us that our problem could be tackled by means of
"intervention analysis" as reported in Box G.E.P. and Tiao G.C.
"Intervention analysis with applications to economic and environmental
problems" (Journal of the American Statistical Association, n. 70,
1975).
Summing up, the actions on the plant should be represented by means of
intervention variables (in the simplest form these could be [0,1]
variables) activated in specific periods and the analysis should point
out the response of the process to such inputs.
This is my first non trivial analysis in the field of time series
and I
never used Stata for this kind of problems: this is the reason for my
question.
Even a quick search of www.scirus.com reveals that there is an
extensive literature on intervention analysis (in fact, the first hit
comes from the RePEc archive, well known to Stata users). Scirus also
suggests narrowing the search using 'arima model' or 'arima models'
which is a good pointer to the statistical method you may need.
HOWEVER, and this is a *big* however, before you ask how to do this
in Stata you probably need to read up on the method, decide whether
it is the best suited method for your needs (which means identifying
alternatives) and, finally, figure out how to get Stata to do it for
you.
The plus is that there is a lot written on it, and it's very
accessible. It's hard to beat Scirus for searches like this one.
Ron�n Conroy
[email protected]
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
