[Clyde Slick]

On 22 Ian, 02:18, Oliver Costich wrote:


I smell a gold ear.

Prove it!!!!
..

[Clyde Slick]

On 22 Ian, 06:57, "Arny Krueger" wrote:


Following borglet's thinking - we should all sell everything we own and
spend it all on a wild night in Law Vegas, because we cannot prove with
absolute certainty that the world will end tomorrow.

Nor can you prove that you will actually have fun in Las Vegaqs

[Clyde Slick]

On 22 Ian, 07:01, "Arny Krueger" wrote:


BTW it was a fun trip,

I won! ;-)-


http://www.golf-products.co.uk/prodi...by%20prize.JPG

[Clyde Slick]

On 22 Ian, 11:07, Oliver Costich wrote:
On Tue, 22 Jan 2008 06:23:01 -0800 (PST), Clyde Slick





wrote:
On 22 Ian, 02:00, Oliver Costich wrote:
On Tue, 22 Jan 2008 03:48:18 GMT, "JBorg, Jr."

wrote:
Oliver Costich wrote:
JBorg, Jr.wrote:

Well now! *Disproving that the sound differences heard by
audiophiles do not physically exist is -- certainty not in the
realm of statistical analysis.

Disproving that the sound differences BELIEVED to beheard by
audiophiles actually exist is not provable or disprovable by
statistical methods [if] your standard is 100% certainty. Nothing is,
other than 1+1=2 and its ilk.

If that is the case, what are the reason(s) you persistently refer to
audiophiles as *golden ear cult*, and why?

Because they always fall back on bull**** like this when they fail to
produce evidence.

That's what the argument is about - some claim to hear things that
allow them the distinguish but can't (at least in this test)
demonstrate it.

But the test did not proved that the subtle difference did not exist.

Of course not absolutely. But then again disproving that something
exists when no one has observed it is pretty hard, like for
leprechauns.

You are not proving whether or not differnces exist.
they may exist for some people, but not exist for others.
we are talking about perceptions.
there is no "THING" to exist, or not exist.

Then these "perceptions" should be good enough to get statistically
valid results. Testing an individual is different than that for a
population.- Ascunde citatul -

- Afiºare text în citat -

The population, or at least most of
the populations, are irrelevant.
As for indiviual perception, for
waqht other consumer preferences do you blind test yourself for
and make statistiacal analyses?
What do you do about choosing Swiss cheese, steak, ice cream, undearm
deoderant, toilet paper,
strawberry jam, automobiles, pencil sharpeners, toasters, your wife?

[Clyde Slick]

On 22 Ian, 18:02, Oliver Costich wrote:


If you are suggesting that the population of concern is not everyone
who can hear, fine. Is it people who listen to music? How narrow do
you want to make it? It depends what you are out to test.


I want it to be people just like me. Identical to me, in every way,
shape and form.




You can restrict the population that way if you choose but then you
can't extend the conclusion of the test to larger ones.


You can't extens the conclusion to anyone,
other than those who took the test.

[Clyde Slick]

On 22 Ian, 18:12, Oliver Costich wrote:



How would you select the 250 people?

I wouldn't.

[George M. Middius]

McInturd said:

If you are suggesting that the population of concern is not everyone
who can hear, fine. Is it people who listen to music? How narrow do
you want to make it? It depends what you are out to test.

[snip]

You can restrict the population that way if you choose but then you
can't extend the conclusion of the test to larger ones. Everyone wants
to eliminate people who firmly believe you cannot dsitiguish between
the cables. You are left with people who believe you can tell and
those that don't know. You could further narrow it to people who
don't know and toss everyone with prejudices.

I nominate Ollie the Collie for this month's RAO Obtuseness Award.

According to Ollie's illogic, haute cuisine should be judged by people who
never dine at fine restaurants. And art should be judged by people who can
barely read their comic books. And jewelry should be judged by those who
never purchase it and never wear it, and fine wine by those who
customarily knock back boilermakers and Thunderbird.

Let's hear it for the uninitiated, says Ollie the Molly, their opinions
are every bit as valuable as people who have spent years appreciating the
best goods on the market.

[George M. Middius]

McInturd said:

How 'borgish of you to excerpt the tiniest, out-of-context rationalization
for your pollution of RAO. Why don't you review the *entire* charter? Get
back to me when you have figured out what "opinion" means.

How is this out of context?

Apparently you don't read very well, or maybe you only read the parts that
appeal to your 'borgish nature.

No "rebuttal" from the statistics-lover?

Evidently your definition of "opinions" excludes subjecting them to
standard scientific method.

BWAHAHAHAHAHA!!!! "Scientific method" on a Usenet chat group! LOLOL!

Sorry. I didn't realize we were limited her to the "pull it out of
your ass" approach.

That's what Normals call an "excluded middle argument". Krooger kalls it
"abuse". Are you proud of yourself for abusing the Krooborg?

In seriousness, the notion that statistical prediction is part of the
scientific method used by real scientists is new to me. Did you misspeak,
or is a huge leap of logic invisible to me?

[George M. Middius]

Clyde Slick said:

You can restrict the population that way if you choose but then you
can't extend the conclusion of the test to larger ones.

You can't extens the conclusion to anyone,
other than those who took the test.

Another violation of borgma. Are you trying to set off a jihad on RAO?

[Oliver Costich]

On Tue, 22 Jan 2008 15:26:04 -0800 (PST), Clyde Slick
wrote:

On 22 Ian, 11:07, Oliver Costich wrote:
On Tue, 22 Jan 2008 06:23:01 -0800 (PST), Clyde Slick





wrote:
On 22 Ian, 02:00, Oliver Costich wrote:
On Tue, 22 Jan 2008 03:48:18 GMT, "JBorg, Jr."

wrote:
Oliver Costich wrote:
JBorg, Jr.wrote:

Well now! *Disproving that the sound differences heard by
audiophiles do not physically exist is -- certainty not in the
realm of statistical analysis.

Disproving that the sound differences BELIEVED to beheard by
audiophiles actually exist is not provable or disprovable by
statistical methods [if] your standard is 100% certainty. Nothing is,
other than 1+1=2 and its ilk.

If that is the case, what are the reason(s) you persistently refer to
audiophiles as *golden ear cult*, and why?

Because they always fall back on bull**** like this when they fail to
produce evidence.

That's what the argument is about - some claim to hear things that
allow them the distinguish but can't (at least in this test)
demonstrate it.

But the test did not proved that the subtle difference did not exist.

Of course not absolutely. But then again disproving that something
exists when no one has observed it is pretty hard, like for
leprechauns.

You are not proving whether or not differnces exist.
they may exist for some people, but not exist for others.
we are talking about perceptions.
there is no "THING" to exist, or not exist.

Then these "perceptions" should be good enough to get statistically
valid results. Testing an individual is different than that for a
population.- Ascunde citatul -

- Afiºare text în citat -

The population, or at least most of
the populations, are irrelevant.
As for indiviual perception, for
waqht other consumer preferences do you blind test yourself for
and make statistiacal analyses?
What do you do about choosing Swiss cheese, steak, ice cream, undearm
deoderant, toilet paper,
strawberry jam, automobiles, pencil sharpeners, toasters, your wife?


You are missing the point. This was about a tset that purported to
show something.

[Oliver Costich]

On Tue, 22 Jan 2008 18:52:37 -0500, George M. Middius cmndr _ george
@ comcast . net wrote:



McInturd said:

If you are suggesting that the population of concern is not everyone
who can hear, fine. Is it people who listen to music? How narrow do
you want to make it? It depends what you are out to test.

[snip]

You can restrict the population that way if you choose but then you
can't extend the conclusion of the test to larger ones. Everyone wants
to eliminate people who firmly believe you cannot dsitiguish between
the cables. You are left with people who believe you can tell and
those that don't know. You could further narrow it to people who
don't know and toss everyone with prejudices.

I nominate Ollie the Collie for this month's RAO Obtuseness Award.

According to Ollie's illogic, haute cuisine should be judged by people who
never dine at fine restaurants. And art should be judged by people who can
barely read their comic books. And jewelry should be judged by those who
never purchase it and never wear it, and fine wine by those who
customarily knock back boilermakers and Thunderbird.

Let's hear it for the uninitiated, says Ollie the Molly, their opinions
are every bit as valuable as people who have spent years appreciating the
best goods on the market.



Point out exactly where I said that.

[Oliver Costich]

On Tue, 22 Jan 2008 18:55:46 -0500, George M. Middius cmndr _ george
@ comcast . net wrote:



McInturd said:

How 'borgish of you to excerpt the tiniest, out-of-context rationalization
for your pollution of RAO. Why don't you review the *entire* charter? Get
back to me when you have figured out what "opinion" means.

How is this out of context?

Apparently you don't read very well, or maybe you only read the parts that
appeal to your 'borgish nature.

No "rebuttal" from the statistics-lover?

Have you got a better tool for testing the abilitiy to discern
differences? The one in your pants isn't big enough.

Evidently your definition of "opinions" excludes subjecting them to
standard scientific method.

BWAHAHAHAHAHA!!!! "Scientific method" on a Usenet chat group! LOLOL!

Sorry. I didn't realize we were limited her to the "pull it out of
your ass" approach.

That's what Normals call an "excluded middle argument". Krooger kalls it
"abuse". Are you proud of yourself for abusing the Krooborg?

In seriousness, the notion that statistical prediction is part of the
scientific method used by real scientists is new to me. Did you misspeak,
or is a huge leap of logic invisible to me?


OK, you tell me what they use. And it's not prediction, but then
apparently the only subtleties you can discern are in cables.

Evidently, much is invisible to you.

[George M. Middius]

McInturd lied:

You are missing the point. This was about a tset that purported to
show something.

That is a lie, you liar. The "tset" in question was the one John Atkinson
described, and all he claimed it showed was the answers given by the
"tsetees".

Why don't you get it through your thick skull that nobody on RAO will ever
make a purchasing decision based on somebody else's "tset"?

[George M. Middius]

McInturd said:

If you are suggesting that the population of concern is not everyone
who can hear, fine. Is it people who listen to music? How narrow do
you want to make it? It depends what you are out to test.

[snip]

You can restrict the population that way if you choose but then you
can't extend the conclusion of the test to larger ones. Everyone wants
to eliminate people who firmly believe you cannot dsitiguish between
the cables. You are left with people who believe you can tell and
those that don't know. You could further narrow it to people who
don't know and toss everyone with prejudices.

I nominate Ollie the Collie for this month's RAO Obtuseness Award.

According to Ollie's illogic, haute cuisine should be judged by people who
never dine at fine restaurants. And art should be judged by people who can
barely read their comic books. And jewelry should be judged by those who
never purchase it and never wear it, and fine wine by those who
customarily knock back boilermakers and Thunderbird.

Let's hear it for the uninitiated, says Ollie the Molly, their opinions
are every bit as valuable as people who have spent years appreciating the
best goods on the market.

Point out exactly where I said that.

I nominate Molly Ollie for the RAO First Quarter Obtuseness Award.

[Andy C[_2_]]

On Wed, 16 Jan 2008 10:52:40 -0800 (PST), John Atkinson wrote:

Okay, I was reading the article and noticed some strange things. The
article says the following:

"I set up a room with two sound systems, identical except for one
component. Everything except the speakers was hidden behind screens."

So he is saying that there were actually two separate systems - two source
components, two amplifiers, etc. But were there two different sets of
speakers too? One would hope not! Using a single set of speakers, there
would need to be a switching arrangement to switch the speakers between the
outputs of the two different amplifiers through the two different speaker
cables. But if there were a properly designed switching network, there
would be no need for two different systems at all. There could just be a
transfer switch using the highest quality relays to switch between the two
speaker cables. That is, a two-throw at the amplifier end and a two-throw
at the speaker end of each speaker cable. This would hold everything else
constant. If there were really two different sets of speakers, then the
experiment was so poorly designed it isn't even worth discussing. Just the
speaker position difference alone would likely cause differences in the
sound that would be measureably far greater than any cable could cause.

Then it also says:

"Using two identical CD players, I tested a $2,000, eight-foot pair of
Sigma Retro Gold cables from Monster Cable, which are as thick as your
thumb, against 14-gauge, hardware-store speaker cable."

Two identical CD players and what else? This guy is being very vague. I
guess he is just addressing the typical WSJ reader who isn't familiar with
or does not care about this stuff. There just isn't enough info provided
to evaluate whether the test setup is valid or not.

[Clyde Slick]

On 22 Ian, 18:57, George M. Middius cmndr _ george @ comcast . net
wrote:
Clyde Slick said:

You can restrict the population that way if you choose but then you
can't extend the conclusion of the test to larger ones.
You can't extens the conclusion to anyone,
other than those who took the test.

Another violation of borgma. Are you trying to set off a jihad on RAO?

only if i get the virgins, and FIRST, before
I detonate. I don't trust God. He plays lots of tricks on us.

[George M. Middius]

McInturd said:

Apparently you don't read very well, or maybe you only read the parts that
appeal to your 'borgish nature.

No "rebuttal" from the statistics-lover?

Have you got a better tool for testing the abilitiy to discern
differences?

Yes, I have my ears. I guess you don't understand that because you've had
yours replaced with 'borg implants.

The one in your pants isn't big enough.

Hey look -- the 'borg tried to make a joke. Ollie, I think your "joke"
rightfully belongs in the Krooborg's millennial collection of artifacts.

Evidently your definition of "opinions" excludes subjecting them to
standard scientific method.

BWAHAHAHAHAHA!!!! "Scientific method" on a Usenet chat group! LOLOL!

Sorry. I didn't realize we were limited her to the "pull it out of
your ass" approach.

That's what Normals call an "excluded middle argument". Krooger kalls it
"abuse". Are you proud of yourself for abusing the Krooborg?

Glad to see you've given up the "science" nonsense on Usenet.

In seriousness, the notion that statistical prediction is part of the
scientific method used by real scientists is new to me. Did you misspeak,
or is a huge leap of logic invisible to me?

OK, you tell me what they use. And it's not prediction, but then
apparently the only subtleties you can discern are in cables.

You're babbling, Ollie. You ask what "they" (presumably real scientists")
use .... but for what? My point is that the scientific method is used by
real scientists for primary research, not for predicting the likelihood of
horse races or consumer audio evaluations.

Evidently, much is invisible to you.

Your qualifications to babble about statistics on RAO are invisible, for
starters. OTOH, I can see your warped ideological handicap quite clearly.

[Clyde Slick]

On 22 Ian, 20:32, Oliver Costich wrote:
On Tue, 22 Jan 2008 15:26:04 -0800 (PST), Clyde Slick





wrote:
On 22 Ian, 11:07, Oliver Costich wrote:
On Tue, 22 Jan 2008 06:23:01 -0800 (PST), Clyde Slick

wrote:
On 22 Ian, 02:00, Oliver Costich wrote:
On Tue, 22 Jan 2008 03:48:18 GMT, "JBorg, Jr."

wrote:
Oliver Costich wrote:
JBorg, Jr.wrote:

Well now! *Disproving that the sound differences heard by
audiophiles do not physically exist is -- certainty not in the
realm of statistical analysis.

Disproving that the sound differences BELIEVED to beheard by
audiophiles actually exist is not provable or disprovable by
statistical methods [if] your standard is 100% certainty. Nothing is,
other than 1+1=2 and its ilk.

If that is the case, what are the reason(s) you persistently refer to
audiophiles as *golden ear cult*, and why?

Because they always fall back on bull**** like this when they fail to
produce evidence.

That's what the argument is about - some claim to hear things that
allow them the distinguish but can't (at least in this test)
demonstrate it.

But the test did not proved that the subtle difference did not exist.

[Clyde Slick]

On 22 Ian, 20:37, Oliver Costich wrote:

George

Have you got a better tool for testing the abilitiy to discern
differences? The one in your pants isn't big enough.

He uses the two tools that bookend his brain.
BTW, they are huge, he is related to
H. Ross Perot.

http://distractiblemind.ambulatoryco...ear%5B3%5D.jpg




OK, you tell me what they use. And it's not prediction, but then
apparently the only subtleties you can discern are in cables.

Evidently, much is invisible to you.-


So much for the benefits of
repeated eye gougings.

[Clyde Slick]

On 22 Ian, 22:33, "ScottW" wrote:
"Clyde Slick" wrote in message

...

On 22 Ian, 18:02, Oliver Costich wrote:

If you are suggesting that the population of concern is not everyone
who can hear, fine. Is it people who listen to music? How narrow do
you want to make it? It depends what you are out to test.

I want it to be people just like me. Identical to me, in every way,
shape and form.

*I'm having problems with that vision......


I wouldn't let my clones in my house, either.
But you can eye gouge them and test them all
you want, in a faraway lab in Newfoundland.

[Shhhh! I'm Listening to Reason!]

On Jan 22, 5:02*pm, Oliver Costich wrote:
On Tue, 22 Jan 2008 13:57:41 -0800 (PST), "Shhhh! I'm Listening to

Reason!" wrote:
On Jan 21, 8:51*pm, Oliver Costich wrote:
On Mon, 21 Jan 2008 13:01:21 -0800 (PST), "Shhhh! I'm Listening to
Reason!" wrote:

I've taken statistics.

I think a true "random" population is counterproductive for perception
tests, as I said. In a true random sample of which painting someone
preferred, I'd expect the distribution of the random population sample
to approximate the percentages of colorblind, or totally blind, people
found in the general population, for example. One or two of that
sample may even know something about art.

First, for testing hypotheses, a random sample isn't enough. It needs
to be a SIMPLE random sample. There is a difference between a random
population (whatever that is, but I get the gist) and a random sample
from a population. That a population need not include everyone you
should have learned from that course.

It's been a while. But no matter.

What would you use to test perceptions?

People who have been trained to detect differences. Everybody I've
seen talk about audio testing states that that detecting subtle
differences is a difficult task. I conclude from that that untrained
participants are likely not to succeed.

I'm not suggesting that this was the case here, or relating this in
any way to the WSJ article. I'm just curious. It seems to me that for
issues of perception a truly "random" population is counterproductive.

It is unless you are looking to home in on the truth. I don't know of
any statistical method for drawing conclusions about population
parameters from sample statistics that doesn't require that samples be
simple random samples. Randomness alone is not enough. It has to be
simple random which in this particualr case means that every group of
39 has an equally likely chance of being selected. One of the problems
with this test is that the "respondents" were self-selected or
otherwise not randomly selected.

You are going down a road I just specifically excluded. Why?

Because a "random population" is a not a term used in statistics.

I meant the part where I said, "I'm not suggesting that this was the
case here, or relating this in any way to the WSJ article." I asked a
general question and specifically excluded it in terms of this test.

I thought you were worried about accurate communication.

Populations are the whole collection of entities for which you want to
test (or estimate) a parameter. Samples can be random but populations
can't. It's not clear what you are talking about.

If you are suggesting that the population of concern is not everyone
who can hear, fine. Is it people who listen to music? How narrow do
you want to make it? It depends what you are out to test.

I want to test whether differences exist between cable. People who do
audio tests professionally state that the listener has to be trained
in order to give a valid test. Therefore, any test that someone
conducts using untrained participants which shows "random guessing" is
suspect to me, as that is the result that I'd expect out of such a
group.

That is why my line of questioning started out with "I wonder how
those most highly qualified fared as opposed to the population tested
in general". Would someone either highly experienced or trained to
detect subtle differences fare better than any person off the street
in detecting, say, a difference between 1% and 5% of some form of
amplifier distortion?

My hypothesis would be "yes" it is likely they would be.

It's like taking a poll on the death
penalty by asking people who walk by your front door. If the test was
sponsored by anyone who has an interest in speaker cable differences
being heard, then agoin the test is suspect. Virtually every
elementary statisitics text gives similare examples of faulty data
collection.

Tell that to the opponents of global warming here. They do not
understand that. One of those people is even now claiming "proofs" in
this very thread, Isn't that ironic?

I understand that. Critical listening is not something people are born
with. Arny, for example, has stated that several times. So have
several others who are actually involved in audio testing. So you
necessarily have to select from a group of those who are interested in
the thing being tested if you use audio or some other related area of
perception as an example.

You can restrict the population that way if you choose but then you
can't extend the conclusion of the test to larger ones. Everyone wants
to eliminate people who firmly believe you cannot dsitiguish between
the cables. You are left with people who believe you can tell and
those that don't know. *You could further narrow it to people who
don't know and toss everyone with prejudices.

As long as they're trained in detecting these differences I don't
particularly care. If they don't care it seems to me it would be a
waste of time.

On the other hand maybe you just want to make the population thsoe who
claim that they can choose the more expensive cable. Is that the one
we're interested in? BTW, does anyone know how the sample at CES was
selected?

I only know what was in the article.

[JBorg, Jr.[_2_]]

George M. Middius wrote:
McInturd said:






If you are suggesting that the population of concern is not everyone
who can hear, fine. Is it people who listen to music? How narrow do
you want to make it? It depends what you are out to test.

[snip]

You can restrict the population that way if you choose but then you
can't extend the conclusion of the test to larger ones. Everyone
wants to eliminate people who firmly believe you cannot dsitiguish
between the cables. You are left with people who believe you can
tell and those that don't know. You could further narrow it to
people who don't know and toss everyone with prejudices.

I nominate Ollie the Collie for this month's RAO Obtuseness Award.

According to Ollie's illogic, haute cuisine should be judged by
people who never dine at fine restaurants. And art should be judged
by people who can barely read their comic books. And jewelry should
be judged by those who never purchase it and never wear it, and fine
wine by those who customarily knock back boilermakers and Thunderbird.

Let's hear it for the uninitiated, says Ollie the Molly, their
opinions are every bit as valuable as people who have spent years
appreciating the best goods on the market.





Holly Molly !
say's Ollie Collie

I'm hopelessly lonely !
since they frontal loboto me

They filled my head with guacamole
because they know that I'm a loony-toony !


LoT"S

[JBorg, Jr.[_2_]]

Oliver Costich wrote:
JBorg, Jr. wrote:
Oliver Costich wrote:
JBorg, Jr. wrote:
Oliver Costich wrote:
JBorg, Jr. wrote:
Oliver Costich wrote:
Mr.clydeslick wrote:
Oliver Costich wrote:




snip

Back to reality: 61% correct in one experiment fails to reject
that they can't tell the difference. If the claim is that
listeners can tell the better cable more the half the time,
then to support that you have to be able to reject that the
in the population of all audio interested listeners, the
correct guesses occur half the time or less. 61% of 39
doesn't do it. (Null hypothesis is p=.5, alternative
hypothesis is p.5. The null hypthesis cannot be rejected
with the sample data given.)

In other words, that 61% of a sample of 39 got the correct
result isn't sufficient evidence that in the general
population of listeners more than half can pick the better
cable.

So, I'd say "that's hardly that".

you seem to be mixing difference with preference, you reference
both, for the same test.

For the purpose of statistical analysis it makes no difference.

But for the purpose of sensible analysis, shouldn't it makes a
difference.

I don't think so. I can't see any way the statistical analysis
would be different.

Preferences are, statistically, immeasureable if the claim is that
listeners can tell the better cable more the half the time.


Agree or Disagree ?


I have no idea what you are asking.


You are admitting that, for the purpose of statistical analysis,
it would make no difference whether the participant determine
or discern subtle differences based on sound differences
or sound preferences during audio testing.


Look! If they could discern these differences then they would make
correct choices. Since enough didn't make correct choices, you have no
support for the existence of the subtle differences



LoL!


What happen if enough participants refuse to make correct choices
because the sound supplied during testing didn't suit their taste ?





Mr. Costich, do you still meant to say that mixing differences with
preferences during testing would make no difference for the purpose
statistical analysis ?

What the hell are you trying to ask?


Yes or No ?




snip

[Clyde Slick]

On 22 Ian, 23:16, "Shhhh! I'm Listening to Reason!"
wrote:


What would you use to test perceptions?

People who have been trained to detect differences. Everybody I've
seen talk about audio testing states that that detecting subtle
differences is a difficult task. I conclude from that that untrained
participants are likely not to succeed.


I must say this. If the differences are that subtle that one has
to be specially trained to hear them, and work at trying to hear them,
then to me,
the differences would be so slight as not to warrant much attention,
time,
effort or cost in my trying to attain the optimal equipment.

[Clyde Slick]

On 22 Ian, 23:50, "JBorg, Jr." wrote:



What happen if enough participants refuse to make correct choices
because the sound supplied during testing didn't suit their taste ?


they are not even given the option to provide aq correct respone,
in such cases where they cannot discern a difference.

[JBorg, Jr.[_2_]]

Clyde Slick wrote:
JBorg, Jr. wrote:



You are admitting that, for the purpose of statistical analysis,
it would make no difference whether the participant determine
or discern subtle differences based on sound differences
or sound preferences during audio testing.

Mr. Costich, do you still meant to say that mixing differences with
preferences during testing would make no difference for the purpose
statistical analysis ?

Yes or No ?

Hehehe, to be fair, give him the option to answer
"I don;t know"!!!!


LoL !


aBig Timeout: my newsreader has just drop 90%
of all posting content from Rao... and other ng.
I may have to go to google to post if I could
remember how... has this just happen to
anyone ?

[JBorg, Jr.[_2_]]

Clyde Slick wrote:
JBorg, Jr. wrote:



What happen if enough participants refuse to make correct choices
because the sound supplied during testing didn't suit their taste ?


they are not even given the option to provide aq correct respone,
in such cases where they cannot discern a difference.



Right. What if the participant got tired of ogling for differences
and decide to just listen to the sound they prefer and forgot
they were taking a test after a while.

[dizzy]

Shhhh! I'm Listening to Reason! wrote:

On Jan 22, 6:01*am, "Arny Krueger" wrote:
"dizzy" wrote in message




Arny Krueger wrote:

I responded to this claim:
"Not that can be retrieved using the search engine at
www.aes.org, Mr. Krueger, using all the alternative
spellings
of your name, and searching both the index of published
papers
and the preprint index. Could you supply the references,
please."
Mr Atkinson seems to have me confused with his research
department. I guess economic cut-backs have affected the
staffing at Stereophile and instead of relying on paid
staff, Mr Atkinson has been forced to go begging for
help on Usenet. :-(

That's quite the illogical (and snotty) remark, Arny.

How snotty and hypocritical of you dizzy!

Why would you say that?

Am I "snotty and hypocritical" for asking you to back up a claim you
made?

You're a sick individual.

Very disappointing, but not entirely surprising, that Arny would
immediately start lying.

[Clyde Slick]

On 23 Ian, 03:57, dizzy wrote:


Very disappointing, but not entirely surprising, that Arny would
immediately start lying.-

Hmmm, that assumes that he took a
pause foro his previous lying.

Prove it!

[Arny Krueger]

"Oliver Costich" wrote in
message
On Tue, 22 Jan 2008 18:52:37 -0500, George M. Middius
cmndr _ george @ comcast . net wrote:



McInturd said:

If you are suggesting that the population of concern is
not everyone who can hear, fine. Is it people who
listen to music? How narrow do you want to make it? It
depends what you are out to test.

[snip]

You can restrict the population that way if you choose
but then you can't extend the conclusion of the test to
larger ones. Everyone wants to eliminate people who
firmly believe you cannot dsitiguish between the
cables. You are left with people who believe you can
tell and those that don't know. You could further
narrow it to people who don't know and toss everyone
with prejudices.

I nominate Ollie the Collie for this month's RAO
Obtuseness Award.

According to Ollie's illogic, haute cuisine should be
judged by people who never dine at fine restaurants. And
art should be judged by people who can barely read their
comic books. And jewelry should be judged by those who
never purchase it and never wear it, and fine wine by
those who customarily knock back boilermakers and
Thunderbird.

Let's hear it for the uninitiated, says Ollie the Molly,
their opinions are every bit as valuable as people who
have spent years appreciating the best goods on the
market.

Point out exactly where I said that.

The Middiot lies profusely, just like borglet.

It's ironic that they should hasten to call other people liars, when they
are habitual liars. Actually it all fits - they are usually lieing when they
call other people liars.

[Arny Krueger]

"Andy C" wrote in message


Two identical CD players and what else? This guy is
being very vague. I guess he is just addressing the
typical WSJ reader who isn't familiar with or does not
care about this stuff. There just isn't enough info
provided to evaluate whether the test setup is valid or
not.

Andy your analysis is good, but you don't have to look that hard to see how
the alleged test is invalid. It was single blind when it could have been
double blind with very little additional expense in terms of time or money.
Single blind tests don't control as many important relevant variables as
double blind tests - the test was too simple and shoddily done to be worth
much analysis.

[Oliver Costich]

On Tue, 22 Jan 2008 19:10:20 -0800 (PST), Clyde Slick
wrote:

On 22 Ian, 20:32, Oliver Costich wrote:
On Tue, 22 Jan 2008 15:26:04 -0800 (PST), Clyde Slick





wrote:
On 22 Ian, 11:07, Oliver Costich wrote:
On Tue, 22 Jan 2008 06:23:01 -0800 (PST), Clyde Slick

wrote:
On 22 Ian, 02:00, Oliver Costich wrote:
On Tue, 22 Jan 2008 03:48:18 GMT, "JBorg, Jr."

wrote:
Oliver Costich wrote:
JBorg, Jr.wrote:

Well now! *Disproving that the sound differences heard by
audiophiles do not physically exist is -- certainty not in the
realm of statistical analysis.

Disproving that the sound differences BELIEVED to beheard by
audiophiles actually exist is not provable or disprovable by
statistical methods [if] your standard is 100% certainty. Nothing is,
other than 1+1=2 and its ilk.

If that is the case, what are the reason(s) you persistently refer to
audiophiles as *golden ear cult*, and why?

Because they always fall back on bull**** like this when they fail to
produce evidence.

That's what the argument is about - some claim to hear things that
allow them the distinguish but can't (at least in this test)
demonstrate it.

But the test did not proved that the subtle difference did not exist.

Of course not absolutely. But then again disproving that something
exists when no one has observed it is pretty hard, like for
leprechauns.

You are not proving whether or not differnces exist.
they may exist for some people, but not exist for others.
we are talking about perceptions.
there is no "THING" to exist, or not exist.

Then these "perceptions" should be good enough to get statistically
valid results. Testing an individual is different than that for a
population.- Ascunde citatul -

- Afiºare text în citat -

The population, or at least most of
the populations, are irrelevant.
As for indiviual perception, for
waqht other consumer preferences do you blind test yourself for
and make statistiacal analyses?
What do you do about choosing Swiss cheese, steak, ice cream, undearm
deoderant, toilet paper,
strawberry jam, automobiles, pencil sharpeners, toasters, your wife?

You are missing the point. This was about a tset that purported to
show something.-

we can agree on that, its like religionsits
trying to prove the existence of god and atheists
trying to prove that God doesn't exist.
It's all about individual faith, you can't
prove whether or not any differences exist.

You are arguing over a useles and irrelevant point.

I pointed out before that believing that people can distinguish the
cables is parallel to religion because the the data don't pan out for
the believers. If it did, then those who insist you can't tell would
become the believers in their own position.

[Oliver Costich]

On Tue, 22 Jan 2008 20:16:53 -0800 (PST), "Shhhh! I'm Listening to
Reason!" wrote:

On Jan 22, 5:02*pm, Oliver Costich wrote:
On Tue, 22 Jan 2008 13:57:41 -0800 (PST), "Shhhh! I'm Listening to

Reason!" wrote:
On Jan 21, 8:51*pm, Oliver Costich wrote:
On Mon, 21 Jan 2008 13:01:21 -0800 (PST), "Shhhh! I'm Listening to
Reason!" wrote:

I've taken statistics.

I think a true "random" population is counterproductive for perception
tests, as I said. In a true random sample of which painting someone
preferred, I'd expect the distribution of the random population sample
to approximate the percentages of colorblind, or totally blind, people
found in the general population, for example. One or two of that
sample may even know something about art.

First, for testing hypotheses, a random sample isn't enough. It needs
to be a SIMPLE random sample. There is a difference between a random
population (whatever that is, but I get the gist) and a random sample
from a population. That a population need not include everyone you
should have learned from that course.

It's been a while. But no matter.

What would you use to test perceptions?

People who have been trained to detect differences. Everybody I've
seen talk about audio testing states that that detecting subtle
differences is a difficult task. I conclude from that that untrained
participants are likely not to succeed.

Then test from the population of the people who are sufficiently
trained, whatever that is. I wouldn't want, howver, for it to be done,
not get statistical significance for the claim and then get the
argument that you need even better training.


I'm not suggesting that this was the case here, or relating this in
any way to the WSJ article. I'm just curious. It seems to me that for
issues of perception a truly "random" population is counterproductive.

It is unless you are looking to home in on the truth. I don't know of
any statistical method for drawing conclusions about population
parameters from sample statistics that doesn't require that samples be
simple random samples. Randomness alone is not enough. It has to be
simple random which in this particualr case means that every group of
39 has an equally likely chance of being selected. One of the problems
with this test is that the "respondents" were self-selected or
otherwise not randomly selected.

You are going down a road I just specifically excluded. Why?

Because a "random population" is a not a term used in statistics.

I meant the part where I said, "I'm not suggesting that this was the
case here, or relating this in any way to the WSJ article." I asked a
general question and specifically excluded it in terms of this test.

I thought you were worried about accurate communication.

I do, but I'm not gertting much of it.

Populations are the whole collection of entities for which you want to
test (or estimate) a parameter. Samples can be random but populations
can't. It's not clear what you are talking about.

If you are suggesting that the population of concern is not everyone
who can hear, fine. Is it people who listen to music? How narrow do
you want to make it? It depends what you are out to test.

I want to test whether differences exist between cable. People who do
audio tests professionally state that the listener has to be trained
in order to give a valid test. Therefore, any test that someone
conducts using untrained participants which shows "random guessing" is
suspect to me, as that is the result that I'd expect out of such a
group.

Suppose you can train people to make consistent selections. The next
question is does it make any real difference in the enjoyment of
listening to music outside of testing. I have never denied that there
are differences but many can be explained by properties of the cable
like resistance, capacitance, etc.

I know an individual that can discern very subtled changes from
switching out capacitors or reversing polarity on a CD player, but
audio design is what he does for a living.

That some people can be trained to do well in tests says little about
what the differences mean to the average listener or even the average
'audiophile'.

So the issue becomes one of what population you want to make the claim
about, or whether you want to make the claim about a specific
individual, though that John can consistently identify things means
little to the rest of us.


That is why my line of questioning started out with "I wonder how
those most highly qualified fared as opposed to the population tested
in general". Would someone either highly experienced or trained to
detect subtle differences fare better than any person off the street
in detecting, say, a difference between 1% and 5% of some form of
amplifier distortion?

My hypothesis would be "yes" it is likely they would be.

Then test that hypothesis. Define the population of highly qualified,
take a simple random sample of a decent size and do the analysis. You
have to specify the population up front becasue you can't define the
population as those who do well on the test. You need to specify the
sample size up front to prevent testing until you get enough positives
to support the claim.

It's like taking a poll on the death
penalty by asking people who walk by your front door. If the test was
sponsored by anyone who has an interest in speaker cable differences
being heard, then agoin the test is suspect. Virtually every
elementary statisitics text gives similare examples of faulty data
collection.

Tell that to the opponents of global warming here. They do not
understand that. One of those people is even now claiming "proofs" in
this very thread, Isn't that ironic?

I understand that. Critical listening is not something people are born
with. Arny, for example, has stated that several times. So have
several others who are actually involved in audio testing. So you
necessarily have to select from a group of those who are interested in
the thing being tested if you use audio or some other related area of
perception as an example.

You can restrict the population that way if you choose but then you
can't extend the conclusion of the test to larger ones. Everyone wants
to eliminate people who firmly believe you cannot dsitiguish between
the cables. You are left with people who believe you can tell and
those that don't know. *You could further narrow it to people who
don't know and toss everyone with prejudices.

As long as they're trained in detecting these differences I don't
particularly care. If they don't care it seems to me it would be a
waste of time.

On the other hand maybe you just want to make the population thsoe who
claim that they can choose the more expensive cable. Is that the one
we're interested in? BTW, does anyone know how the sample at CES was
selected?

I only know what was in the article.

[Oliver Costich]

On Tue, 22 Jan 2008 21:10:17 -0800 (PST), Clyde Slick
wrote:

On 22 Ian, 23:16, "Shhhh! I'm Listening to Reason!"
wrote:


What would you use to test perceptions?

People who have been trained to detect differences. Everybody I've
seen talk about audio testing states that that detecting subtle
differences is a difficult task. I conclude from that that untrained
participants are likely not to succeed.


I must say this. If the differences are that subtle that one has
to be specially trained to hear them, and work at trying to hear them,
then to me,
the differences would be so slight as not to warrant much attention,
time,
effort or cost in my trying to attain the optimal equipment.


If you have the time and enjoy doing it, why not? The silliness comes
in with the cables in the thousands of dollars range. In particular
cables are an area of the system where you get the least bang for the
buck (at retail - have you seen the obscene dealer margins of the
exotic cables?)

[Oliver Costich]

On Wed, 23 Jan 2008 02:58:10 GMT, Andy C wrote:

On Wed, 16 Jan 2008 10:52:40 -0800 (PST), John Atkinson wrote:

Okay, I was reading the article and noticed some strange things. The
article says the following:

"I set up a room with two sound systems, identical except for one
component. Everything except the speakers was hidden behind screens."

So he is saying that there were actually two separate systems - two source
components, two amplifiers, etc. But were there two different sets of
speakers too? One would hope not! Using a single set of speakers, there
would need to be a switching arrangement to switch the speakers between the
outputs of the two different amplifiers through the two different speaker
cables. But if there were a properly designed switching network, there
would be no need for two different systems at all. There could just be a
transfer switch using the highest quality relays to switch between the two
speaker cables. That is, a two-throw at the amplifier end and a two-throw
at the speaker end of each speaker cable. This would hold everything else
constant. If there were really two different sets of speakers, then the
experiment was so poorly designed it isn't even worth discussing. Just the
speaker position difference alone would likely cause differences in the
sound that would be measureably far greater than any cable could cause.

Then it also says:

"Using two identical CD players, I tested a $2,000, eight-foot pair of
Sigma Retro Gold cables from Monster Cable, which are as thick as your
thumb, against 14-gauge, hardware-store speaker cable."

Two identical CD players and what else? This guy is being very vague. I
guess he is just addressing the typical WSJ reader who isn't familiar with
or does not care about this stuff. There just isn't enough info provided
to evaluate whether the test setup is valid or not.


All good points. This particualr test was badly enough designed to be
flawed from the start, never mind what the data actually conclude.

[Oliver Costich]

On Tue, 22 Jan 2008 20:50:31 -0800, "JBorg, Jr."
wrote:

Oliver Costich wrote:
JBorg, Jr. wrote:
Oliver Costich wrote:
JBorg, Jr. wrote:
Oliver Costich wrote:
JBorg, Jr. wrote:
Oliver Costich wrote:
Mr.clydeslick wrote:
Oliver Costich wrote:




snip

Back to reality: 61% correct in one experiment fails to reject
that they can't tell the difference. If the claim is that
listeners can tell the better cable more the half the time,
then to support that you have to be able to reject that the
in the population of all audio interested listeners, the
correct guesses occur half the time or less. 61% of 39
doesn't do it. (Null hypothesis is p=.5, alternative
hypothesis is p.5. The null hypthesis cannot be rejected
with the sample data given.)

In other words, that 61% of a sample of 39 got the correct
result isn't sufficient evidence that in the general
population of listeners more than half can pick the better
cable.

So, I'd say "that's hardly that".

you seem to be mixing difference with preference, you reference
both, for the same test.

For the purpose of statistical analysis it makes no difference.

But for the purpose of sensible analysis, shouldn't it makes a
difference.

I don't think so. I can't see any way the statistical analysis
would be different.

Preferences are, statistically, immeasureable if the claim is that
listeners can tell the better cable more the half the time.


Agree or Disagree ?


I have no idea what you are asking.


You are admitting that, for the purpose of statistical analysis,
it would make no difference whether the participant determine
or discern subtle differences based on sound differences
or sound preferences during audio testing.


Look! If they could discern these differences then they would make
correct choices. Since enough didn't make correct choices, you have no
support for the existence of the subtle differences



LoL!


What happen if enough participants refuse to make correct choices
because the sound supplied during testing didn't suit their taste ?


How much more absurd can this get?



Mr. Costich, do you still meant to say that mixing differences with
preferences during testing would make no difference for the purpose
statistical analysis ?

What the hell are you trying to ask?


Yes or No ?




snip

[Oliver Costich]

On Tue, 22 Jan 2008 21:11:46 -0800 (PST), Clyde Slick
wrote:

On 22 Ian, 23:50, "JBorg, Jr." wrote:



What happen if enough participants refuse to make correct choices
because the sound supplied during testing didn't suit their taste ?


they are not even given the option to provide aq correct respone,
in such cases where they cannot discern a difference.


Use a different test in which the outcomes are "I hear a difference"
and "I don't hear a difference". This becomes a two-tailed hypothesis
test which requires a higher proportion (than a choose the better one
test) of people to detect a differnce to support that they can tell.

[Oliver Costich]

On Tue, 22 Jan 2008 22:04:29 -0800, "JBorg, Jr."
wrote:

Clyde Slick wrote:
JBorg, Jr. wrote:



What happen if enough participants refuse to make correct choices
because the sound supplied during testing didn't suit their taste ?


they are not even given the option to provide aq correct respone,
in such cases where they cannot discern a difference.



Right. What if the participant got tired of ogling for differences
and decide to just listen to the sound they prefer and forgot
they were taking a test after a while.


Yes indeed it can get more absurd.

[Oliver Costich]

On Tue, 22 Jan 2008 22:07:28 -0500, George M. Middius cmndr _ george
@ comcast . net wrote:



McInturd said:

Apparently you don't read very well, or maybe you only read the parts that
appeal to your 'borgish nature.

No "rebuttal" from the statistics-lover?

Have you got a better tool for testing the abilitiy to discern
differences?

Yes, I have my ears. I guess you don't understand that because you've had
yours replaced with 'borg implants.

The one in your pants isn't big enough.

Hey look -- the 'borg tried to make a joke. Ollie, I think your "joke"
rightfully belongs in the Krooborg's millennial collection of artifacts.

Evidently your definition of "opinions" excludes subjecting them to
standard scientific method.

BWAHAHAHAHAHA!!!! "Scientific method" on a Usenet chat group! LOLOL!

Sorry. I didn't realize we were limited her to the "pull it out of
your ass" approach.

That's what Normals call an "excluded middle argument". Krooger kalls it
"abuse". Are you proud of yourself for abusing the Krooborg?

Glad to see you've given up the "science" nonsense on Usenet.

In seriousness, the notion that statistical prediction is part of the
scientific method used by real scientists is new to me. Did you misspeak,
or is a huge leap of logic invisible to me?

OK, you tell me what they use. And it's not prediction, but then
apparently the only subtleties you can discern are in cables.

You're babbling, Ollie. You ask what "they" (presumably real scientists")
use .... but for what? My point is that the scientific method is used by
real scientists for primary research, not for predicting the likelihood of
horse races or consumer audio evaluations.

Evidently, much is invisible to you.

Your qualifications to babble about statistics on RAO are invisible, for
starters. OTOH, I can see your warped ideological handicap quite clearly.



Thanks for the free psychoanalysis. It''s worth even less that I paid
for it.

Your knowledge of what real scientists do using statistics is
underwhelming.

[Arny Krueger]

"Oliver Costich" wrote in
message

Your knowledge of what real scientists do using
statistics is underwhelming.

The real person behind the Middiot persona knows far more than he lets on.
This particular persona is all about ridicule. Unless you're particularly
fond of being ridiculed, don't bother.