rdf said:
OMG, so now there are measurements that affect sound? That sounds like something a stone deaf, graph staring meter reader might say. But it's great news!!
So now you can not only list your test methods for establishing audible differences, you can include the necessary measurable performance parameters of the system. Perhaps you can "reveal" the secret that Andre refused to divulge? The world awaits....
cheers,
AJ
Bonsai said:
Nope. There is no reason that the guy swapping the cables over has to be blinded as to which cable he is attaching. Single-blind (just the listener) is statistically perfectly adequate.
I am not going to build ABX equipment just to achieve DBT status if it is no more discriminating than single-blind.
I'm sure your psychologist wife is aware of the current concerns over the validity of DBTs in medical circles. Things like exclusion criteria can seriously distort the results yet the fact that it is a DBT gives a piece of research this legendary "scientifically respected methodology" status and almost guaranteed publishing. A double edged sword! Ever wondered how 2 high quality DBTs can come to opposite conclusions?
Jakob2,
I think you misunderstand - I am the only listener here. It is multiple tests just with me. So a lot of variables are removed from the equation. And I am not in the least offended!
Cheers,
Alan
SY said:
Wow - coincidence - I told my daughter the story earlier this evening!
I've published my planned methodology above. My daughter will do the cable swapping - unblinded. But she and I will not see each other nor speak to each other at any point during the entire test. Perhaps that technically makes it a DBT - it certainly addresses the "Clever Hans" type concerns.
rdf said:
In a real scientific test, it is not an option.
No one here is interested in tests that prove nothing.
I am sure there are examples of faulty tests in all fields but the methodology of DBT has surely been nailed down by people who know what they are doing.
The results are a hard pill to swallow, as what they prove goes against what we experience.
fredex said:
I ask for a third time. After we get proof that we can not distinguish cables after being subjected to DBT, will this prohibit the 'illusions' of sighted listening to reappear? (Later in normal -real life- listening.) If faith dies, miracles stop I think. But if miracles continue? Maybe sighted listening is far more analytical because one sense helps the other, and we must just delete fallibility pitfalls, but under sighted conditions? Maybe by a panel of people listening alone, keeping notes, and not allowed to communicate until results are processed?
Your input please.
@ Alan Hope,
Yes, i realized that. I only wanted to point out, that among the roughly 100 different participants on my tests no one was performing as well as usual before doing some training under the specific test protocol.
Some test protocols were apparently easier to get used to as other, but even under these training was needed.
The only exception from this rule i´ve found so far, was in the case that listeners didn´t know they were participating in a somewhat ´official´ test procedure, because the protocol did follow extremely close their normal way while trying out new equipment.
But this is at this time just an informal statement as it has to be confirmed by future tests.
That´s why i emphasize strongly the use of controls; positive controls give you some certainty that the required sensitivity level was reached under the specific test conditions and the negative control assures against the ´clever hans´
@ Salas,
It´s only anecdotical evidence (as it was reported by members on several forums) that they cured themselves from being ´golden ear´ after participating in blind tests and failing.
After that they were convinced that for example cables doesn´t make an audible difference, and that they before were influenced by autosuggestion and argued that they now didn´t hear any differences anymore and if they would hear any differences in the future they now know that is would be only suggestion.
Others, who failed in dbts, said they heared the differences in normal listening situations again, and remained using the ´voodoo stuff´ .
You are absolutely right in stating that `one sense helps the other´ because we normally are used to live in that way. I´d think this is well known and confirmed in the field of psychology.
That is one reason why people need some training for blind tests.
Jakob2
Yes, i realized that. I only wanted to point out, that among the roughly 100 different participants on my tests no one was performing as well as usual before doing some training under the specific test protocol.
Some test protocols were apparently easier to get used to as other, but even under these training was needed.
The only exception from this rule i´ve found so far, was in the case that listeners didn´t know they were participating in a somewhat ´official´ test procedure, because the protocol did follow extremely close their normal way while trying out new equipment.
But this is at this time just an informal statement as it has to be confirmed by future tests.
That´s why i emphasize strongly the use of controls; positive controls give you some certainty that the required sensitivity level was reached under the specific test conditions and the negative control assures against the ´clever hans´
@ Salas,
It´s only anecdotical evidence (as it was reported by members on several forums) that they cured themselves from being ´golden ear´ after participating in blind tests and failing.
After that they were convinced that for example cables doesn´t make an audible difference, and that they before were influenced by autosuggestion and argued that they now didn´t hear any differences anymore and if they would hear any differences in the future they now know that is would be only suggestion.
Others, who failed in dbts, said they heared the differences in normal listening situations again, and remained using the ´voodoo stuff´ .
You are absolutely right in stating that `one sense helps the other´ because we normally are used to live in that way. I´d think this is well known and confirmed in the field of psychology.
That is one reason why people need some training for blind tests.
Jakob2
SY said:
Strawman! Shame on you SY. Where did I question or attack his abilities to design and measure? I have no clue what he designs and would probably have little or no interest unless it was loudspeakers - which I doubt.
I am asking him (or any other DBT rejectionist) point blank to explain what valid, scientific method(s) he uses to determine audible differences and what measurable parameters are important for such a system of stellar resolution. I (and others) can only benefit from such information....assuming it is applicable in the world of physical reality. If it is "something" outside of the sound waves impinging upon the ears and occurring only in the recipient advocates head, then it will do me no good. But I know that is not what he or similar claimer is advocating. So I await...
cheers,
AJ
salas said:
Firstly I like Jakob2 's reply, especially
This has been my experience... I have learnt to be more objective by a better understanding of how my senses operate. But like most DIYers I like building stuff not participating in endless Blind Testing so I am still a victim of auto suggestion. If I pickup a component and it is Mil spec or used by NASA I will probably use it just because I have "faith" that it will sound better even though it may not.
As to being "cured" by DBT I guess as Jakob2 says it depends on the individual.
fredex said:
Thank you.
Would you consider as controlled testing the following idea?:
Some sighted component evaluation of say 5 listeners taking turns.They will be not allowed any communication, and they will have a specific form to keep notes. Time will be individually slack enough. The data will be processed by a 6th person who never listened to the component.
salas
Maybe you would need another "control" group who were sent identical amps that were actually the same. But they would be told that the same story as the first group.
My guess is that the control group would hear differences because they believed that the amps were different. Interesting idea.
Maybe you would need another "control" group who were sent identical amps that were actually the same. But they would be told that the same story as the first group.
My guess is that the control group would hear differences because they believed that the amps were different. Interesting idea.
fredex said:
We will probably get high fallibility amongst the absolutely identical amps group.
We humans are fallible. No doubt about it.
My attention would be mostly in the actual written sonic descriptions amongst the first group. The group that knows there are cheap or expensive components in two identically looking sealed amps.
Hello all, Excuse me for my late night bloviations, but Iv'e been drinking high octane Russian beer, so here goes! This pontification apply's to subjectivists and objectivists in equal measure...... To be a good scientist you must be willing to eat your own children! When you come up with a hypothesis no matter how elegant or appealing, you must immediately attack it as if it were your worst enemy. It is only through this trial by fire that truth is separated from dross. This rigor also keeps you open minded as you admit that all ideas including your own are human and therefor fallible. The path of shurety is the path of the fool. Yoda, eat your heart out!
- Status
- Not open for further replies.