Neo Sunrider wrote:
> I am just taking an undergraduate introductory stats course but now I
> am faced with a somewhat difficult problem (at least for me).
>
> If I want to test a hypothesis (t-test, z-score etc.) and the underlying
> distribution will under no circumstances approach normal... (i.e. the results
> of the experiment will always be something like 100*10.5, 40*-5 etc.) The
> Central Limit Theorem doesn't help here, or does it?
>
> Can anyone explain, or point me in the right direction - how can I test in
> these cases?
>
> Thanks alot
>
> A.
>
Test if the distribution is nowhere 'near normal': generate a whole pot
load of data (at least 100 points), form nominally identical
conditions. Make a histogram. Eyeball it. Then run some of the
various Normal testing tests on it. If your eye can't say it is not, it
is probably close enough for a 't' test.
Second, the Central Limit Theorem says the distribution of the
_contrast_ (i.e., the average of a sample) tends toward a normal
distribution. It says nothing about the individual observations. Only
certain very special distributions that do not converge properly are not
included, and they may fall under the CLT's power, too.
If your dist. is clearly not normal, but looks log normal or the like,
consider a transformation on the raw data.
--
Jay Warner
Principal Scientist
Warner Consulting, Inc.
4444 North Green Bay Road
Racine, WI 53404-1216
USA
Ph: (262) 634-9100
FAX: (262) 681-1133
email: [EMAIL PROTECTED]
web: http://www.a2q.com
The A2Q Method (tm) -- What do you want to improve today?
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
