Jerry Bromenshenk wrote:
>I'd have to go digging, but I can tell you that as a general rule, the
>power of the test improves through a sample of 25 (bees, light bulbs,
>etc.). Then you hit the old issue of diminishing returns. The amount of
>improvement in the power or reliability of the test begins to fall off
>rapidly between 25 and 30, and sample sizes over 30 don't add much for the
>time invested. Now, this generalization assumes lots of things about
>normal distributions, representative samples, and other issues - so it
>won't hold for all estimates -- but its not a bad rule of thumb.
Jerry's memory serves him well. Norman Bailey covered that point in
Chapter 5 of his book, STATISTICAL METHODS IN BIOLOGY; he wrote (p. 34):
"[Statistical test] results are all right provided [sample size] is
sufficiently large, say greater than 30."
*********
We often have a hard time, intuitively, accepting such notions.
However, intuition doesn't always provide the best perspective. For
instance, here is an exercise you can conduct and amaze your friends:
If more than people are in a room, you can announce (claim you are
psychic, or whatever) that two of them have their birthdays on the same day
of the year.
Start by asking all who have birthdays in January and see if any match
up, etc.
I did that exercise to relieve the boredom at a couple's club party some
years ago (about three dozen people present). It turned out that two pairs
had their birthday on the same day of the year; one of that pair even had
their birthday on the day of the party! The party livened up after that!
Adrian
Adrian M. Wenner (805) 963-8508 (home phone)
967 Garcia Road (805) 893-8062 (UCSB FAX)
Santa Barbara, CA 93106 [http://www.beesource.com/pov/wenner/index.htm]
*****************************************************
*
* "We not only believe what we see:
* to some extent we see what we believe."
*
* Richard Gregory (1970)
*
*****************************************************
|
