In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Alan McLean) writes:
>[EMAIL PROTECTED] wrote:
>>
>> More importantly, I would say: DON'T DO TESTS. Instead, try to find
>> models that you would be prepared to use to predict the response
>> in as-yet untried circumstances.
>> --
>
>Hypothesis testing is simply one useful method of identifying 'models
>that you would be prepared to use to predict the response
> in as-yet untried circumstances.'
>...
Yes. Mea culpa. Of course I myself do tests (not least informal ones
during model-building), but I became carried away while responding after
a hard day at the office.
I've just given lectures including what I would call two-tailed
chi^2 and F tests, analysing well-known data from Georg Mendel
and from Cyril Burt, and resulting in very small values of the
corresponding test statistics, strongly suggesting fiddled data
[more accurately, strongly suggesting analysing further data sets
to see if the evidence of fiddling/massaging becomes overhelming].
Are some people using "two tailed" to mean something other than
looking for extreme values in both tails of the test statistic?
-- Ewart Shaw
--
J.E.H.Shaw [Ewart Shaw] [EMAIL PROTECTED] TEL: +44 2476 523069
Department of Statistics, University of Warwick, Coventry CV4 7AL, U.K.
http://www.warwick.ac.uk/statsdept/Staff/JEHS/
yacc - the piece of code that understandeth all parsing
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
