on 30/05/2003 10:23 am, R. Allan Reese at [email protected] wrote:
> Dynamic graphics are
> like a random walk: you may find a pretty view, but does that make it
> meaningful? If there is some objective measure to optimize, should that
> not be built into the program? If not, it is a design decision.
A walk is never random when you bring your ideas with you. It's more like
the process of diagnosis: you start with ideas that you bring to the
situation, but you need to make sure the situation is not more complicated
than your ideas. In most cases where it has made a difference, EDA has
refined models which I would have drawn from the previous literature, or, in
at least one case, suggested they were plain wrong.
But it brings up a whole nightmare in the philosophy of science. P-values
for parameter estimates are contingent on the model being specified in
advance. If I derive my model (and not just estimate its parameters) from my
data, the usual logic of hypothesis testing is violated (and this,
incidentally, is the biggest objection to stepwise methods: their p-values
are not p-values).
Whether you would shed tears over the loss of p-values depends on yourself,
of course.
Ronan M Conroy ([email protected])
Lecturer in Biostatistics
Royal College of Surgeons
Dublin 2, Ireland
+353 1 402 2431 (fax 2764)
--------------------
--------------------------------------------------------------------------------------------------------------------
This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom
they are addressed.
If you have received this email in error please notify the
originator of the message. This footer also confirms that this
email message has been scanned for the presence of computer viruses.
Any views expressed in this message are those of the individual
sender, except where the sender specifies and with authority,
states them to be the views of The Royal College Of Surgeons in Ireland.
--------------------------------------------------------------------------------------------------------------------
*
* 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/
