A terrific paper along these lines, which probably should be required
reading for all psychology graduate students is Meehl (1967). I've
reproduced the link below. ....Scott
http://www.tc.umn.edu/~pemeehl/074TheoryTestingParadox.pdf
Ken Steele wrote:
I don't know if you have ever had the odd experience of telling a
student that the outcome of some statistical test means that the null
hypothesis can be rejected and having the student look at you with
trepidation and asking "Is that good"?
I now have had an alice-in-wonderland experience of having my daughter
explain to me how statistical tests are described in her intro
statistics course as taught in a math department. (I checked her
textbook to make sure that her story was accurate.)
Psychology Statistics: Rejecting the Null Hypothesis is GOOD.
We perform a simple experiment by collecting a bunch of people and
randomly assign them to one of two groups. One group is exposed to a
treatment condition and the other is not. We hypothesize that the
treatment condition is efficacious in changing behavior but we don't
know. The null hypothesis is that the treatment had no effect. So we
calculate some test statistic to see if there is a difference. Finding
a difference is good. It means that we can write up the report and
send it off to a journal. Failing to reject means that we did a lot of
work but it is unlikely that the report will appear in print.
Math Statistics: Rejecting the Null Hypothesis is BAD.
Here is what my daughter is being taught.
I have a machine that make dinguses. A dingus must be 2 inches long
(5.08 cm for our Euro friends). A 2-inch dingus is a good dingus. Any
systematic deviation from 2 inches is bad. So I perform an
"experiment" by taking a random sample of dinguses and calculating
their length. My null hypothesis (Ho) is that my sample comes from a
population that has a specific mean value (2 inches). The alternative
hyppothesis (Ha) is that my sample comes from a population that does
not have that value (mu <> 2 inches). It is good to fail to reject the
null because that means your machine is working correctly. Rejecting
the null is bad because it means that your machine is on the fritz and
you must go out and hire some expensive labor to repair the machine.
(And they are probably unionized or employ undocumented workers or
some other horror that management would like to avoid.)
Here is what my daughter copied from the blackboard in her statistics
class, "The null is what you want to happen."
Ken
---------------------------------------------------------------
Kenneth M. Steele, Ph.D. [EMAIL PROTECTED]
Department of Psychology http://www.psych.appstate.edu
Appalachian State University
Boone, NC 28608
USA
---------------------------------------------------------------
---
You are currently subscribed to tips as: [EMAIL PROTECTED]
To unsubscribe send a blank email to
[EMAIL PROTECTED]
--
Scott O. Lilienfeld, Ph.D.
Associate Professor
Department of Psychology, Room 206
Emory University
532 N. Kilgo Circle
Atlanta, Georgia 30322
(404) 727-1125 (phone)
(404) 727-0372 (FAX)
Home Page: http://www.emory.edu/PSYCH/Faculty/lilienfeld.html
The Scientific Review of Mental Health Practice:
www.srmhp.org
The Master in the Art of Living makes little distinction between his work and
his play, his labor and his leisure, his mind and his body, his education and
his recreation, his love and his intellectual passions. He hardly knows which
is which. He simply pursues his vision of excellence in whatever he does,
leaving others to decide whether he is working or playing. To him – he is
always doing both.
- Zen Buddhist text
(slightly modified)
---
You are currently subscribed to tips as: [email protected]
To unsubscribe send a blank email to [EMAIL PROTECTED]
