----- Original Message -----
From: DeLa <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Wednesday, March 22, 2000 6:27 AM
Subject: Sample size: way tooo big?
> I have been trying to explain to some co-workers that a sample
> can be too big.
> That is not very easy because it is contratictory to what
> intuition says.
>
> Can someone point me to some good arguments or literature?
> Or correct me if my assumption is wrong?
I can see that:
(1) a sample can be bigger than is justified in terms of cost
(2) a sample can be bigger than ethically justified, if subjects are
inconvenienced, harmed, etc.
(3) a sample can be bigger than is easy to work with, (especially if you
are using a limited demo version of a package!)
(4) a sample can be too big for certain implicit assumptions about the
relationship between practical and statistical significance to be
valid;
(5) a sample can be too big for certain types of graphical
epresentation, designed for small-to-midsized data sets to work
well - eg, a boxplot of 10E6 points will show thousands of outliers
even with a normal distribution;
(6) a sample can be too big for certain "pathological" tests. For
instance, Tukey's "7-11" test, applied to huge samples, will almost
certainly yield results based entirely on some slight contamination
of one or the other distribution that would be irrelevant to a
normal-sized sample.
(7) in certain testing procedures, equal sample size is assumed. This
may be either an essential assumption or simply a matter of
robustness. In either case, one sample may be too big for the rest.
But, in general, as Tom T. Hall might have put it, "It's faster horses;
older whiskey; younger [members of the appropriate sex]; and MORE DATA!"
-Robert Dawson
===========================================================================
This list is open to everyone. Occasionally, less thoughtful
people send inappropriate messages. Please DO NOT COMPLAIN TO
THE POSTMASTER about these messages because the postmaster has no
way of controlling them, and excessive complaints will result in
termination of the list.
For information about this list, including information about the
problem of inappropriate messages and information about how to
unsubscribe, please see the web page at
http://jse.stat.ncsu.edu/
===========================================================================
