Cable Directionality (Moved Threadjacking)

[SY]

'Scuse me, Peter, what did you say?

[Peter Daniel]

I'm here to please only

[fdegrove]

Hi,

Cheers,

[Ophth]

....for someone who hears the difference, it's not a belief anymore...

I hate to say this because it may sound insulting, but...
Some of us wonder if that someone only believes that they hear a difference. We wonder if there really is no audible difference. If that speculation is correct, then the difference he believes he hears is a result of psych, placebo effect, or whatever.

I think that is the reason that some want to actually do a blind study. So that the hearers of the difference can prove that there is a difference. Maybe then we doubters will stop bugging them all the time, and join them in discussing why it happens and how to make things better.

So - back to sample size. Could everyone - on both "sides" tell me just how much they expect to hear?
What % of time will you say that the "bad" cables are bad?
What % of time will you say that the "good" cables are bad?
That way we can establish a sample size for this study.

[Peter Daniel]

Originally posted by Ophth

I hate to say this because it may sound insulting, but...
Some of us wonder if that someone only believes that they hear a difference. We wonder if there really is no audible difference.

I hate to say it, but if you substitude a word wonder with believe, you won't be much off the mark either.

[Ophth]

...if you substitude a word wonder with believe, you won't be much off the mark...

OK, so we're agreed that we each hold a somewhat tentative belief that is so far unproven?

[Peter Daniel]

I agree with that, but OTOH, I can function perfectly well without a proof.

I don't really care if something is proven or not. Proof is only needed by people who are not certain about things. I understand (or rather "feel") very well the world around me and I live in perfect harmony with it

[Ophth]

Proof is only needed by people who are not certain about things.

Says it all.
What about those etimates?

[ThatGuy]

In order to make this test simpler, and yeild results tht are valid, I believe that a change to the methdology presented so far is needed. First, I would take advantage of system that is alreday accepted as valid by researchers in sufficient trial numbers: ABX. It would be simple to emulate this: Create 30 identical stereo interconnects. Divide this into a subgroup of 10 sets(Of each subgroup 1 is X, 2 is A, 3 is A). The test subject will attempt to identify in each subgroup, whetehr X is the same as A or B. 10 subgroups represents 10 trials. 10 trials is the general accepted minimum number of trials that has a statistically significant value. Each interconnect should have a direction marked. The direction is not true direction, nescarrily, but an established radom direction to test the interconnect established by a random generator or coin tosses, etc., and recorded by the maker of the test samples for later reference when scoring. Ir ealize that this may seem excessive(the number of test samples), but I feel it is important to use an accepted form of protocol. IF, indeed, the cables are cosmetically identical, and no breach of the ID is revealed, and the directon can not be measured, then IMO the compiled test result data should provide very valuable evidence into whether cable directionality is audible.

-ThatGuy

[Ophth]

What you suggest would be a great way to set up the test.
There's no problem with their study design, though.

They've already established what they're looking for subjectively.
Messed up imaging.
They can do that with just the two pairs of wires.

I would recommend that the coding for each cable pair be hidden. That way each person can do multiple tests. Of course, no one should know if set A or B is correct until the very end. But they should also re-hide the "A" and "B" for each test.

But however the test is set up, first it must be established what % difference in correct "poor image" results between A and B we are looking for.