Re: Probability problem

*** Date: Nov 24, 2007 5:17 PM
Author: Jack Tomsky
Subject: Re: Probability problem
Since the population mean and standard deviation are
given, he does have 100% confidence intervals on the
population mean, the population standard deviation,
the probability that a single time reading will be
within specified limits and the probability that the
average of ten time readings will be within specified
limits. Jack***

MY COMMENT

STUPIDITY IN ACTION

Jack Tomsky does not understand (YET) that this kind
of problems (absolutely usual in terms of Statistical
learning and training)
His claim
*** he does have 100% confidence intervals in the
population mean, the population standard deviation…
***
is an ABSOLUTE ABSURDITY in Statistical Frequencist
terms.
IN FACT
In this kind of problem the Population are supposed
fixed however UNKNOWN
Consequently:
Each sample data WILL PROVIDE A (DIFFERENT)
) CONFIDENCE INTERVAL for each (of the two)
parameter’s values: different but all, by
construction, with a fixed probability (v.g. 95%) to
contain the respective parameter.
Facing the very deficient John Tomsky´s education on
Statistics (for example concerning the fact that
Critical Values are obtainable by Monte Carlo
Simulations, and that Null Hypotheses cannot be
proved true) I’m not surprised for his insistence to
display nonsense.
****
Since I authorised a paper on Critical Values
throughout Monte Carlo I invite (IT`S THE SECOND
TIME)John Tomsky to write to
Revue de Statistique Appliquee (he can write in
English, of couse) claiming I’m wrong: very, very
curious about this refereed Journal response.
The paper is L. Amaral Afonso, P. Duarte *** Un
Nouveau Test pour la Distribution Uniforme***
(1992, XXXX (1), 77-79)
*************

Luis Amaral Afonso



This is another example of why no one should ever use Afonso statistics. Despite several years of everyone trying to explain to him the difference between test statistics and parameters, he is incapable of understanding the difference.

When the population mean is specified as 100 seconds, a 100% confidence interval for the population mean is [100, 100].

Jack
.

Loading