Re: Hypothesis Testing: the TEST STATISTIC

Reef Fish wrote:
Russell.Martin@xxxxxxx wrote:
Reef Fish wrote:
Russell.Martin@xxxxxxx wrote:
Reef Fish wrote:
Keywords: Hypothesis Testing; Test Statistic; Type I error;
alpha=Pr(Type I error); p-value.

This is a mini-lecture of an important topic at the Freshman Level.
The lecture is prompted by the conspicuous silence of both the
pros in industry and those who teach, at Afonso's latest blunders
about the TEST STATISTIC in the problem of testing the
equality of two proportions (a topic found in most first couses
in statistics).

If others are like me, they've tuned out Afonso.

I knew that. But as I said later, his SUBJECT headlines about
Bob Ling might have at least aroused some interesting in finding
out what the fuss was about.

I didn't see those as new threads, and if they were under
an old thread I missed them because the thread had reached
the point where I had dropped following it. That may have be
the case with others.

There aren't that many Afonso threads. Each of them lasts a
LONG time, because they not only fragments into little pieces,
Afonso would sometimes inject comments in a completely
unrelated thread to introduce more of his Bob Ling subjects.
The frequency of appearance of Bob Ling subjects by Afonso
is a good indication that it is the SAME thread.

And even if they WERE his derogatory labels for different
threads, wouldn't one be inclined to tell him to SHUT UP, or
or succinctly, STFU?

Clearly just a waste of bandwidth to try, so I don't.
YMMV.



The pros in industry seldom deal with this GENERAL topic of
Hypothesis Testing. They just use the values of the test statistcs
spewed out by SPSS or SAS or the p-values printed by the
software packages associated with the test statistics.

Several of those who are known professors and often speak
up against Afonso's errors, or posted about OTHER statistical
topics are conspicuously silent in seeing the same error made
by Afonso repeatedly (about a dozen times in one day, amidst
subject lines like "HORRIBILIS Bob Ling" , "The complete
Bob Ling K.O.", and the latest, "REQUIEM for a Bob Ling
BLUNDER". Even if they don't normally read anything posted
by Afonso, those libelous subjects posted by Afonso surely
should have attracted some attention, out of curiosity if
nothing else.

At least two of the professors' silence (Rubin and Tomsky) on
Afonso's blunder was probably because they are so far up the
stratoshperes of academia that they have never taught that
lowly topic [testing the equality of two proportions] (usually
at the Freshman level introductory course). The other
professors are probably not even aware of the issue (or the
topic), or not sure what the fuss is about. Else they surely
could have pointed Afonso to a chapter or section of a
textbook that deals with the particular topic.

You'll have to do intellectual battle with him on your
own, like Roland at the bridge.

Sorry Russell. This is NOT a case of intellectual battle.
The intellectual content of one of the battlers is zero.

A one sided battle still distracts the more powerful side.

You spoke before you read the following paragraph.

I did read the whole post first, but wanted to put that
part of my reply where I did. Sorry you didn't like that
choice. Now I suppose you'll proceed to lecture me
on English composition. Save your time, you're
overreacting.

Perhaps you
should learn to read the entire post before making an out of context
comment.

No, perhaps you shouldn't make assumptions about
how people will choose to respond. :-)

It looks particularly FOOLISH when the NEXT
paragraph says it's NOT a battle of any kind because there was
NO battle to be fought.

That's your position. As I stated, I disagree. Or
maybe it isn't a battle, just a patrol. But that's
how battles can start. Roland was just covering
the retreat of his king's army. Anyway, the entire
analogy is just that, an analogy, and nothing to
get worked up about.


It's a case of protecting the innocent from being dragged
down the gutter by the statistical quacks and the ignoramus,
like Afonso. In fact, the OP even thanked him for all his
bum advice in the 2nd post of the 40 post-thread now.

I applaud your worthy goal, but protecting the gullible
from bad Internet information is a Sisyphusian task.
There are too many gullible people and too much bad
information.

That doesn't preclude one from doing SOMETHING about it.
Richard Ulrich has been giving BAD, BAD, BAD information
in this group ever since he started 10 years ago.

Like I said, I applaud your goal. I'm suggesting that
you pace yourself, in part by not attacking me for
my suggestions, which are meant to be helpful.


It's people like yourself that keeps him going. It's people
who not only did NOT contribute to trying to stop abuse,
but in his only appearance, he makes ridiculous excuses
against those who do.

IMO it wouldn't help.


You underestimate the gullibility of the readership of
sci.stat.math in believing what might appear to you or me
to be complete staistical nonsense.

Not really.

I take that to mean the only thing it could possibly mean,
that you didn't think Afonso or Richard Ulrich is guilty of
any statistical crime of Quackery and that you think those
one-time supporters of Richard Ulrich (but have thinned out
to almost extinction) were not GULLIBLE folks?

No, it means I do not underestimate the unfortunate
gullibility of the readership. I also try not to get worked
up over the fact.


In this case, there's more for the general public's
education, on the topic of Hypothesis Testing, in which
Richard Ulrich had already led the field in his mangling
of the terms and meaning of alpha and beta and
probabilties of Type I and Type II errors. We already
had a length thread about p-values in which most of the
discussant seem blissfully ignorant about what it is,
though they use it everyday from SPSS or SAS. Then
we had threads about the general subject of Hypothesis
Testing, and of testing sharp hypotheses, and more and
more statistical heresy emerged.

In short, the readership as a whole, is woefully uneducated
about the basic CONCEPTS behind the subject of
Hypothesis Testing. This thread happens to touch upon
one PRINCIPLE that is usually hidden in other hypothesis
testing problems -- that is, the Hypothesis Testing result
is NON-EQUIVALENT to the corresponding Confidence
Interval interpretation on the same parameter.

It was what I consider to be my public service, to shed some
light on what every FRESHMAN taking a course in statistics
should know, to a newsgroup many of which are professionals
and teachers, who DO NOT those basics; and at the same
time protect those innocent newbies who had the misfortune
to step into the misguided advice from the most ignorant,
and welcome their quackery with open arms.

THAT's the true battlefield, Russell.

Yes, but perhaps you should just prepare a
standard message, sent once a week, telling
all posters to ignore any advice that doesn't
come from a short list, like Prof. Rubin and
yourself. Then you could spend more time
diving. :-)

Smiley notwithstanding, your comment is nothing short of
being completely asinine.

Ah, smiley impediment alert. :-) Well, your wasting
time trying to take me to task when I'm on your side.

Even Richard Ulrich and Afonso
make occasional posts that are not COMPLETELY
erroneous. I DO NOT stereotype anyone. Richard Ulrich
is entitled to his voice of opinion when he is IN LINE with
the subject matter and the subject content.

It's ONLY when these folks shows their MALPRACTICE
and Quackery that I speak out.

Their continued practice of Quackery highlights Quackery
in different subject areas within Statistics.

Like Roland, I'm sure
you will have piles of Afonsos at your feet before you
are overwhelmed, at which time we will come back to
claim your body and write ballads to your heroism. :-)

Again, your smiley is just a betrayal of your own asinine position.
I didn't say anything the first time -- that your remark is NOTHING
but an insult to equate me with whoever that Roland is, who is
obviously not remotely close to ME or to MY CLASS, and use
that as your derogatory reference.

You need to lighten up. Oh and read up a little
on mythic Western heroes and their place in the
culture, especially Roland. It certainly wasn't
meant as an insult but as what such stories have
always been meant to be, tales to entertain and
transmit the ethic of the culture.


Russell, you can STOP that anytime now.

Just quit replying and I'll drop it, too. It takes two
for a discussion or whatever you call this. But I'm
betting you want the last word, and trying to convince
you to drop it is as much a waste of bandwidth as
you trying to convince Afonso of something. Oh
well...


I have not seen YOU contributing ANYTHING to Statistics or
any recent statistical disussion that would make YOU a an
indispensible reader of this group should you disappeared
wituout a trace tomorrow.

True, but OTOH all of us are dispensible.


NO ONE will miss you. Not even your Roland.

In your opinion most of the people here would not
be missed. Fine, that doesn't bother me.


Fortunately I don't know who thie Roland is, and whatever
heroic battles he fought. I have encountered NO BATTLE
so far. Not a single one -- except perhaps my battle with
Herman Rubin about the necessity or even the usefulness
of Measure Theory (of the Loeve and Halmos kind) in
Applied statistic. In that case, after a few rounds of jousting,
we each settled to call it an agreement to disagree.

I don't engage in any discussion (especially any battle, if there
is such) unless I KNOW I have much more ammunition than
the opponent. :-)

When I am finished with his newsgroup, for lack of subjects
I had not already provided information, and plenty of useful
information, I'll find other subjects and groups for my own
interests.

In that respect, whatever you said of Roland cannot possibly
happen to Reef Fish Bob -- mark my word! :-)

Good, I wouldn't want you pierced but a couple dozen
arrows.

The truth finally emerges? The arrows certainly wouldn't come
from YOU.

No, since as I've said I agree with you.

And they certainly wouldn't come from Afonso or
Richard Ulrich. Who do you have in mind for the couple dozen
arrows?

You are taking it entirely too specifically and personally.
I'm just saying that in general I don't want you pierced
by arrows, as happened to Roland when it was entirely
avoidable. Your posts are too instructive to lose.

snipped for brevity


I think more should lend a hand to the eradiction of statistical
pollution, because silence is often believed by the MOST
guilty ones as readers condoning or even supporting their
trash.

-- Reef Fish Bob.

I would, but I don't know enough which is why I'm
here to learn.

Cheers,
Russell

But apparently you're a very BAD learner. And what happened to
those dozens of arrows of your wishful thinking that might be
pierced on me?

You've misunderstood that whole idea, as I've mentioned.
Please lighten up.


Making one post like this is bad enough for a non-learner like
yourself. Making a SECOND one, to make your same support
for the villants of statistical society is NOT going to help you to
learn.

-- Reef Fish Bob.

LOL. Bob, you've totally overreacted to this. You
take every response as an attack rather than just
an exchange.

Cheers,
Russell

.

Quantcast