Re: Bob Ling´s ignorance is clear

We are getting very close to 2007 and you are still pumping bowel gas
into your private intelluctual vacuum.
Try as you will, you cannot get the pressure above zero. OMU


"Luis A. Afonso" wrote:
*** In a tedious, boring, pointless, nitpicking and error-prone thread on the test of two independent proportions, on the Hypothesis of Ho: p1 = p2, whose APPROXIMATE Z test has only ONE correct Test Statistics, which is taught and given in EVERY TEXTBOOK that has ever been written in Statistics on that particular approximate Z test of the EQUALITY of p1 and p2 -- one high-school calibre participant in this group first made every conceivable ERROR in the execution of the test of that hypothesis, and now resorted to arguing that the Z test is less than EXACT, which we knew all along. ***


My response

WHO DENIES THAT THE Z TEST APPEARS ON ALL TEXT-BOOKS?
(This is completely INVALID as an argument AGAINST THE EXACT TEST. To doubt I was not aware of it is a CALUMNIOUS ACT and the note that I made errors in this matter eludes completely the hesitations and errors made by Bob Ling. I had to post


1) Date: May 12, 2006 11:51 AM
Author: "Luis A. Afonso"
Subject: The FOUR COMMANDMENTS about tests
2) Date: Sep 12, 2006 6:32 PM
Author: "Luis A. Afonso"
NO important if the NULL HYPOTHESES is true or not true
3) Date: Oct 24, 2006 10:55 AM
Author: "Luis A. Afonso"
Subject: The TRUTH about Test Hypotheses

*****************
Continuing Bob Ling´s accusatory plan::

But does that absurdity deserve HUNDREDS of posts, over a period of 6 months or more, with Afonso-created SUBJECT titles of at least a few dozen, for the SAME absurd arguments? Within a period of LESS than 2 years in sci.stat.math, Afonso (actually there are several different Afonsos posting by the same name, but that is a different, but related, issue) managed to have posted enough pure NOISE to have risen to the ALL TIME TOP poster list in the tally by Google:
All time top posters (in sci.stat.math)
1665 lic...@xxxxxxxxxxx
1574 Reef Fish
1192 wpi...@xxxxxxxx
1166 hru...@xxxxxxxxxxxxxxxxxxxx
910 rich.ulr...@xxxxxxxxxxx
871 "Reef Fish"
746 wpil...@xxxxxxxx
703 Reef Fish
661 jtom...@xxxxxxxxxxxxx 590 OsherD
Of the 1665 posts by Afonso, at least 1200 of them were on the test of proportions topic, and the matching-birthday problem. The rest were pointless simulation programs in QBASIC and derogatory and perjorative SUBJECTS, such as "*** you Bob".
So much for the introduction of the problem and the historical perspective of Luis A. Afonso.
Old Mac User wrote:
in theCURRENT thread "Comparing fractions (or proportions)" in which the "discussants" and the frequency of their posts are:
Luis A. Afonso 26
Jack Tomsky 12
Old Mac User 9
Reef Fish 2
Bob O'Hara 1
totaling 50 posts,
OMU> I am amazed that you (Luis) continue to post noise at a OMU place where people come to learn something useful OMU The noise is now more than I can bear. I'm leaving you OMU> to your misery. ***
My response
An EXAGERATED or LYING list of my pots with NULL JUSTIFICATION in what I made only noise. All in Bob Ling´s way to *discuss*.
Are there at least a few points that I remember (I am not
1) The *Birthday Matching Problem* which he couldn´t go no further than he picked from the ELEMENTARY text books. I showed that the problem could be EXTENDED to more general problems in different directions.
2) The Exact similitude between Test Hypotheses and Court Trials he claimed was putted KO .by my demonstration that (on contrary of the generality of HT) we can, in the case persecution wrong identification, prove that the INOCENCE (H0) could be stated..
3) He wrongly stated that from Monte Carlo simulation techniques we cannot go further than results that are *petition principles* nothing of new could be attained.!!! I had the opportunity to show how wrong he is by treating several problems, for example finding critical values of statistics impossible to get analytically like the Lilliefors´s test and the .Geary´s on testing normality.
4) I helped to clarify that the two INDEPENDENT binomial samples comparison could not be performed by the Fisher´s Exact Test because the total columns is taken constant which is completely against any logic.
5) Finally I showed throughout several examples the Binomial Test can be performed exactly based on the first principles. Bob Ling stated that *there is only ONE way o perform* that is using the Z test with pooled variance.
_______licas (Luis A.Afonso)

.