On Sat, 30 Dec 2000 12:36:35 GMT, [EMAIL PROTECTED] (Alfred
Breull) wrote:
> On Fri, 29 Dec 2000 15:57:14 -0500, Rich Ulrich <[EMAIL PROTECTED]>
> wrote:
> >Jim's statement is not a "view"; it is 100% correct: there *is* a
> >literature, and that does not depend at all on "your psychological ...
> >background." What makes your part "damnable" is that Jim wrote
> >neatly and intelligibly, whereas your note has parts that are dubious
> >and other parts that I can't even parse.
alf>
> The problem is not the existence of literature, the problem is the
> content.
>
> However, instead of names calling, you may want to
[ snip, ... essentially 'cite the good literature, in great detail' ]
Was that "name calling"? I said that the your hostile criticism of
Jim (which I quoted), lacks apparent (face) validity since it is
illogical and badly written. And, I guess, it is unfriendly and in
bad taste. Someone can disagree if they please.
If you want to see the literature, I suggest that your own research
could start with "computer aided diagnosis." The paradigm that has
worked well is: comparing the correlations between raters (several
skilled humans plus one computer program). This is not my area, but
as I understand it, the computers do well when the criterion can be
narrowly defined. (And, if the criterion won't change.)
Computers do well in predicting who will stay in school, in every
test, but schools would have to keep the algorithm out of the hands of
their applicants. Computers do better than experts in making medical
diagnoses when the correct answer has to be from a narrow set.
But there is one place where formula-decisions are in use:
The banking industry has almost abandoned human judgment, or
so it seems from what I have read.
alf>
> (4) and is 'demonstrably valid' (and we don't speak of 'face
> validity', do we ?).
Ah, I think it is proper to insist on face validity, first.
For various reasons.
alf>
> Otherwise, we could simply agree on the fact that only little has
> changed in publication practices since Meehl's famous text in 1957 on
> the significance of the insignificance -- and on the statement of
> 'some idiot could offer pseudo stat- reasons for something' and, with
> a little luck, it will be even published in a scientific journal,
> couldn't we ?
Meehl never claimed that there were no good findings....
I believe it was Meehl's phrase, "nomological network," which
emphasized the consistency of findings across related areas.
We ought to believe something, in part, because "it fits in."
Here is a problem for us all: Most scientists (2/3? 4/5?) are not
very good, critical readers -- outside of their narrow areas.
Is that because they don't read enough to have wide "areas"?
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
