Dear all,
I'd like to throw my 2 cents in on the hypothesis testing thread, even
though I'm an inexperienced grad student who has only taught stats for a
few semesters.
First, I'd definitely agree with those who have indicated that if we
abandoned conventional hypothesis testing (as we see it in most
introductory texts) with effect sizes or a bayesian approach, then people
would apply these methodologies just as rigidly and foolishly as they
currently do with hypothesis testing. We would simply replace alpha=.05
with some other alleged statistical law.
At this point of my statistical career, I'm still teaching out of
textbooks chosen by others and covering/leaving out topics as determined
by others (with barely enough time to cover what I "need" to cover).
Also, I'm not so bold yet (and maybe will never be) as to feel hypothesis
testing should be abandoned.
but I do try to give my class the sense that while hypothesis testing is
an important and much-used tool of statistics (that can protect us from
making unwarranted conclusions about the effectiveness of some treatment),
that it should not be applied w/o thought
to do this, I like to do an example that comes out to be statistically
significant BUT is obviously not practically significant... for instance,
a paired samples t-test where we measured sales of beer for a month in
several cities, started an advertising campaign, and then measured sales
of beer the following month...the difference is greater than zero...but
the additional sales do NOT cover the additional cost of the advertising
admittedly a contrived example with plenty of side issues to consider
(will the advertising be worthwhile in the long term? are there other
variables besides the advertising that explain why sales went up? Et
cetera...)
My point is that I want to show my class an example where they can see the
pitfalls of making a decision based solely on a p-value. I don't want
them going "Ok, the p-value is .04 in this problem, so I don't reject, no
wait, I reject, I think, Ok, yeah I reject, so whatever the treatment is
must be good."
___________________________________
Christopher Mecklin
Doctoral Student, Department of Applied Statistics
University of Northern Colorado
Greeley, CO 80631
(970) 304-1352 or (970) 351-1684
===========================================================================
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/
===========================================================================
