Re: Richard Ulrich used THREE posts to distort one post in one thread
- From: Richard Ulrich <Rich.Ulrich@xxxxxxxxxxx>
- Date: Sun, 04 Jun 2006 20:39:02 -0400
On 3 Jun 2006 19:34:52 -0700, "Reef Fish"
<Large_Nassau_Grouper@xxxxxxxxx> wrote:
Jens, an obvious (and self-acknowledged) complete novice in
"linear regression" asked some questions about "linear regression".
Or maybe, since he knows so little, he needs to know more
than "linear regression." - From his reply to me, he needs a
lot of help with design.
Recognizing his lack of training and experience on the subject, I
gave him simple, straightforward answers, and pointed him to
the "Applied Linear Models" book by Neter et al.
That was good.
That was when Richard Ulrich took it as his opening to push his
Quackery in regression analysis. I had already responded
IN FULL in the post in the tinyurl, on June 1..
http://tinyurl.com/ozn76
Here is an oddity of Bob's thinking. He should know by
now that various people think differently, and that he, himself,
tends to read badly. However -
He is *offended* when anyone reads a (long) query any
differently that he reads it, in any tit-or-jottle.
He is *offended* if his own comment is not taken as
100% complete.
He is *offended* if his own comment is not taken as
100% correct (despite ongoing disputes).
Those are patterns with Bob that I've noted several times.
They were really big in an earlier regression thread (Sehwail)
where Bob has yet to appreciate the two main problems.
Ulrich continued to harp on his revisionist version -- getting more
and more away from the fact that the case had long been settled,
to Jens's satisfaction on my recommendation of the Neter et al book.
- but he had not mentioned 5 stages with multiple inputs
and multiple outputs. Linear regression might be the least
of his problems.
Given the above, it may be apparent to everyone that Richard
Ulrich was really hitting below the belt as well as trying to push
his own snake oil medicine, when he wrote, on
Bob calls my approach "quackery" when a more fitting insult
would be, "mainstream mumbo-jumbo." It is "really hitting
below the belt," he says, for me to return his compliment and
to help his understanding of English usage (once again). Below,
I took the opportunity to mention that Bob's style and role, not mine,
better merit the term "quack." Or snake oil, I could add today.
Sat, Jun 3 2006 9:59 pm
when he should be answering to my challenge of his ERRORS
of statement in the "coefficient of variation" thread. to write
down the hypotheses to be tested, when he tried his usual
verbal obfuscation.
RU> It seems unprofessional -- to use Bob's term, "quackery" --
RU> for Bob to ignore the fact that there is a large mainstream of
RU> statistical analysts who will tell you to check out your
RU> univariate predictors. And Bob says, "There's nothing to
RU> worry about."
Which is the exactly correct "simple answer" to a complete novice
who can't even write down correctly a simple regression model!
- if that were Jens's only problem, I might agree, if Bob had
tossed in some affirmative *hint* about who looks at
predictors and why. Incomplete answer from Bob.
Turns out to be important. Still would have been worth
mentioning, to elicit more about the problem, even if it hadn't
turned out important.
EVERYONE is invited to read the first FOUR posts in the
"linear regression" thread -- to by the OP Jens, and two responses
by me, and judge for yourself the appropriateness or inappropriate-
ness of my responses.
Appropriate enough to Jens. Incomplete.
Later ones to me: inappropriate.
The last thing anyone needs on any regression and linear models
thread is any commentary by Richard Ulrich, on MY response.
The last thing anyone *wants* to read is one more
half-assed attack by Bob Ling, on Richard Ulrich, or on
anyone else. That's been said by several people, but that's
"water off the duck's back", as they say.
[...]
--
Rich Ulrich, wpilib@xxxxxxxx
http://www.pitt.edu/~wpilib/index.html
.
- Follow-Ups:
- References:
- Prev by Date: Re: Yurra-Bob: a yet not paid debt
- Next by Date: Re: Richard Ulrich used FOUR posts to distort one post in one thread
- Previous by thread: Richard Ulrich used THREE posts to distort one post in one thread
- Next by thread: Re: Richard Ulrich used FOUR posts to distort one post in one thread
- Index(es):
Relevant Pages
|
