m00es's NOISE explained because he is REPEATING the SAME ad infinitum

m00es wrote:
Reef Fish wrote:
m00es snipped my reply about the relevance in my reply to Jack Tomsky
and continue to make noise here.

There was no need to bring up my name in that reply. In fact, your
reply to Jack did not include an ounce of statistical substance.

You made nothing but NOISE for weeks because you did not
understand the concept of the need in VALIDATING the
assumptions behind a statistical procedure such as a Simple
Regression, before any statistical test or inference can be
validly applied!

You are ASHAMED of the name m00es EXPOSED in relation to
the m00es ERRORS in statistical SUBSTANCE of validation.

You snipped what I said in reply to m00es about the relevance in my
reply to Jack Tomsky on a separate statistical topic in which I readily
and immediately acknowledged an error *I* made, posted out by Jack.

m00es> Reef Fish wrote:

after lamenting how some posters in sci.stat.math would NOT
admit their own errors and spend weeks making NOISE trying
to defend or cover up their own errors. I cited specifically the
names of m00es, Illywhacker, Afonso, Bob O'Hara, and Ulrich.

Still can't believe someone like m00es could be THAT obtuse
to issues of APPLIED statistics, even given the fact that he
was obviously trained by mathematicians who would <sic:NOT>
know how to handle any data. But most them would have caught
on, after about a dozen posts that DATA has nothing to do
with the statement of an Ho. :-)

and m00es followed that with

mooes> Actually, you were the one that kept evading and dodging
mooes> the issue, to cover up the fact that you were wrong.

and continued THREE more rounds of PURE NOISE under his
subject of " <insert neutral topic here>" to COVER UP his
own ignorance about the relevance of DATA in validating the
statistical assumptions in a simple regression. I had already
spared him FOUR consecutive posts in his use of <insert
neutral topic here> for m00es's cover up. NO MORE.
Malpractice as bad as m00es's is NOT going to be covered
up that way, especially after I had already given him a chance
to just keep quiet and disappear, but he chose to return.

Since you kept covering up, by continuing to make NOISE,
here's the statistical SUBSTANCE m00es kept evading to try
to cover up his own ignorance about, explicitly stated in
detail, for the benefit of everyone:

=== Statistical SUBSTANCE of VALIDATION enclosed ===

The ASSUMPTIONS behind doing a Simple Regression are
the ERRORS are i.i.d. N(0, sigma^2). which can be broken
down into three multually independent assumptions which can
be checked ONLY by the RESIDUALS obtained AFTER doing
a simple regression (see: iterative process of sponsor and
critic in George Box's "Science and Statistics" JASA 1976):

1. The RESIDUALS are Normally distributed.
2. The RESIDUALS are independent, hence uncorrelated
with the FITTED Y.
3. The RESIDUALS are homoscedastic.

The RESIDUALS depend on the DATA on X and Y used to
fit the simple regression. The residuals DO NOT depend
on the statement of the Ho or Ha to be tested.

m00es argued ad infinitum that the simple regression
Ho and Ha imply something about the OBSERVED data,
and that the simple regression data can be used to test
corr(X,Y) without VALIDATING the Hogg and Craig
ASSUMPTION he brought in himself (about X andY)
which is entirely different from the regression assumptions
(which has NO assumption about X or Y, but only of
the ERRORS in the conditional distribution of Y given X).
Furthermore m00es repeatedly argued that the DATA is
irrelevant when he considered the distribution of various
test statistics which are valid ONLY when the assumptions
are (and having been validated), but not otherwise.

The preceding paragraph stated m00es's ERRORS in
Statistical SUBSTANCE, confirmed and verified a hundred
times in m00es's repeated NOISE about testing rho and
beta1, easily found in everyone of his posts on the subject
in the archives.

=== Statistical SUBSTANCE of VALIDATION enclosed ===

This post will be repeated EVERY time m00es continues his
NOISE without an explicit rebuttal showing how he had not
violated the statistical SUBSTANCE in the enlosed
paragraphs above.

-- Reef Fish Bob.

.

Quantcast