Switching labels does nothing as he is going to email his results, no one needs to see the cables again.
Edit: Oh I see SY has already addressed this
Last edited:
As long as the same cables isn't used over and over for testing 😀, only 'burned-in' and 'new' should be fine.
For the record, Andre, I'd like to express my respect for you for doing this. You, sir, have mucho balls.
jd
But the statistics for the proposed test method are different. The samples choosen wouldn´t be put back in the bowl for the next guess, so the chance for correct guessing varies for every trial.
Wishes
Edit: Ah, i see it´s not any longer to pick 5 out of ten, as he will not know how many cables were burned in. Sorry.
Last edited:
Yes, that was exactly my point- each cable is RANDOM, and independently so. So each guess is a 50/50 shot.
@ scott wurcer
Thanks....
See what you mean! Quite similar to this thread really......plenty of Comet Heads etc.
Just LOL re Polo and affording a horse!!
Thanks....
See what you mean! Quite similar to this thread really......plenty of Comet Heads etc.
Just LOL re Polo and affording a horse!!
Although a reasonable way to structure it, that adds sort order in as a variable. 10 independent events have fewer variables.
Natural distribution of heads in 10 cointosses:-
w
An externally hosted image should be here but it was not working when we last tested it.
w
Last edited:
No, I did not miss that part. Do you honestly think though that he may answer "Nope, can't hear a difference." Thereby canceling the rest of the test, and admitting to his error? (if I read the plan correctly, which apparently I'm having trouble doing)
I would say that outside this forum it will not be taken seriously, unless it can be independently and easily established that scientific method was followed. The 'wiki' idea is good, at least we can continue to refine until we know it is a clearly objective test.
Apparently not. My impression is that there are far too many variables thus far which have not been anticipated, or accounted for. Secondly, there may be issues with statistical viability. (sample and guess sizes not being large enough etc.) Perhaps I should just stick to reading the wiki as it evolves. I'll go back and try reading it all again.
I'll go back and start again, then. Apologies if I have jumped to erroneous conclusions.
I do however think this would be vastly easier if a proven and accepted method were at least used as a blueprint, and then modified, rather than making it up as we all go along. There is too much which can go wrong, and render the test useless.
Yes, the statistics. For example, if I have chosen 4 as (say) "burned in," then I know that only one of the remainder can be "burned in" even if the remainder is 6.
Yes, yes, and no, in that order.
Yes, yes, and no, in that order.
Remember Andre, 10 out of 10 and maybe none of them are broken in. ;-) See how the game changes?
But if it's as audible as you say, it ought to be like picking out red bits of paper from pink ones.
It should be simple for you to do, pretty much error free. -Regardless if it's one, one hundred, or one thousand attempts.
At least that's what I'm inferring you believe you can hear from your previous posts.
@ SY,
statistically it wouldn´t be a problem as you have only to look for the "best case" for guessing. It´s the old lottery wheel game without putting back.
For the best case chance for guessing the correct number is highest, for the other lower. That´s why 4 (from 5 out of ten samples) correct identifications would be sufficient for p<0.05 .
@ John Curl,
for 10 trials 9 correct answers would be sufficient.
If the number of "burned-in" cables is choosen by flipping coins, both zero and ten "burned-in" would be quite unlikely.
As Andre would receive a "burned-in" and a "unburned-in" clearly labeled he will be able to do a paired preference test (A/B comparison).
I´d tend to the original idea (picking 5 out of ten) intuitively but don´t think that the new idea is unfair.
Wishes
statistically it wouldn´t be a problem as you have only to look for the "best case" for guessing. It´s the old lottery wheel game without putting back.
For the best case chance for guessing the correct number is highest, for the other lower. That´s why 4 (from 5 out of ten samples) correct identifications would be sufficient for p<0.05 .
@ John Curl,
for 10 trials 9 correct answers would be sufficient.
If the number of "burned-in" cables is choosen by flipping coins, both zero and ten "burned-in" would be quite unlikely.
As Andre would receive a "burned-in" and a "unburned-in" clearly labeled he will be able to do a paired preference test (A/B comparison).
I´d tend to the original idea (picking 5 out of ten) intuitively but don´t think that the new idea is unfair.
Wishes
No one that I know has said that cable break-in is very audible. It is just sometimes audible with the right audio equipment and the right audio wires. It is like the difference between a good wine and a great wine. To a wino, there will be very little difference. To a wine buff, there may be a great difference, yet the wine buff might be fooled, on occasion, or by direct manipulation.
spoilsport
ya know... you ought to take some of your own medicine and read up on multivariable statistics... you'd realize just how ridiculous your ridicule of others and these efforts looks... not to mention your lame attempts at discrediting others credentials... and the badgering with "read the book" comments... really sad
While I might respect the innovations in your amplification designs over the decades, I certainly see now your contentious and contemptuous demeanor clouds your ability to apply your knowledge to other endeavors...
ya know... you ought to take some of your own medicine and read up on multivariable statistics... you'd realize just how ridiculous your ridicule of others and these efforts looks... not to mention your lame attempts at discrediting others credentials... and the badgering with "read the book" comments... really sad
While I might respect the innovations in your amplification designs over the decades, I certainly see now your contentious and contemptuous demeanor clouds your ability to apply your knowledge to other endeavors...
Well, some of us know how to design audio, and we tend to keep an open mind about things like cable differences. Others, in related fields, might think that they are experts in audio as well, and they claim without even trying, much of the time, that something is impossible.
- Status
- Not open for further replies.