On 13 Jul, 23:03, Jack Tomsky <jtom... at (no spam) ix.netcom.com
wrote:
On Jul 13, 9:52 pm, Jack Tomsky
jtom... at (no spam) ix.netcom.com> wrote:
The team that do NO ACCEPT THE NULL
HYPOTHESES
___Ronald A, Fisher
___Herman Rubi
___Luis Amaral Afonso
preferring to say that (if so)
there is no sufficient evidence to reject H0
if
the
test value lies out the critical region.
Luis Amaral Afonso [the moderator destroyer]
I never heard of Herman Rubi, but here is what
Professor Herman Rubin wrote on November 26,
2005.
http://www.math.yorku.ca/Who/Faculty/Monette/Ed-stat/
0
241.html
"I have yet to see a classical approach which
considers the balance. But if one assumes that
the
decision maker is consistent, then it is an easy
result that action must consider a linear
combination
of the probabilities of the resulting outcomes in
the
various states of nature. Looked at otherwise,
this
corresponds to minimizing the integral of some
consequence function with respect to some measure
on
the states of nature. This is the behavioristic
Bayesian approach. But it is not the
philosophical
one; only the product of loss and prior is even
considered by this. To take a simple example,
suppose
one is testing
that the mean of a normal distribution is
exactly
0. Then the probability of rejecting the null
hypothesis if it is true must be balanced against
an
integral of the probability of accepting this
hypothesis over the non-zero values of the mean
with
respect to some measure. It is not that hard to
proceed from this type of approach."
What is this business about "the probability of
accepting this (null) hypothesis"?
Jack (moderator)
My comment
YES I made an error: the guy is Herman Rubin.
HOWEVER you called Prof. TIAZ when the ecat name
is
TIAGO (de
Oliveiea) THEN you had 1 letter absent (O) and 1
wrong (Z by
G).Conclusion: YOU ARE WORSE THAN I AM.
BY THE WAY : why you don´t ask him if the Null
Hypotheses can be proved tue, then acceptable?
LET US KNOW WHAT HIS OPINION IS!!!!
Luis Amaral Aoso [the moderator destroyer]
The null hypothesis does not have to be proven to
be true to be accepted. Professor Rubin talks about
the probability of accepting the null hypothesis when
it is not true (i.e., non-zero values of the mean in
his example).
Jack- Ocultar texto citado -
- Mostrar texto citado -
YOU ARE WRONG, Jack Tomsky
From Herman Rubin to Richard Ulrich (a few yers ago):
The point null hypothesis is always false. If the
sample size is not too large, it can be a reasonable
approximation, but it is still not the precise
p-value
which gives the consequences of incorrect rejection.
AS YOU EASILY WOULD GOT, IF YOU WAS A LITTLE
INTELLIGENT, Prof. Herman
Rubin disagrees that one can achieve by Hypotheses
Testing the *
miracle * that consists in find a parameter value
exactly.
I repeat : contact him and get what he thinks about.
I DO NEED NOT, IT
IS CLEAR CRYSTAL FOR ME.
Luis Amaral Afonso [the oderator destroyer]
Professor Rubin has said that the null hypothesis can be accepted even if it is false and has a probability of doing so.
Jack (moderator)- Ocultar texto citado -
- Mostrar texto citado -
