Ronald Bloom wrote...
> Herman Rubin <[EMAIL PROTECTED]> wrote:
> > In article <[EMAIL PROTECTED]>,
> > Jan Ljung <[EMAIL PROTECTED]> wrote:
> >>Does anyone recall the complete quote that Tukey once
made
> >>about "...working less on theorems and more on
practical problems"?
> >>At least I believe that it was Tukey.
>
> > It would not surprise me; the abandonment of the use of
> > theory by Tukey was one of the tragedies of statistical
> > development.
>
> > One should work on practical problems, USING THEORY.
> > Even the first statistics courses should be based on
> > probability, not games of chance, and decision making,
> > which is the balancing of consequences.
>
> And how shall we define "probability" in those first
> statistics courses? As an element of a closed, axiomatic
> mathematical system? I think not.
Why not? First introduce discrete probability space. It is
not too awfully difficult to learn what it is. Then explain
how it corresponds to something in real life like dice or
coin. It's better not to mix mathematical objects and
phisical objects.
-----------------------------
Alexander Tsyplakov
Novosibirsk State University
http://www.nsu.ru/ef/tsy/
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
