SLL: Did you read in tonights paper the latest horrible news on this study.

AP, New York - As Yale physicists continued their grueling work on breaking the secret of sub-particle interconnect variance and its' effect on audible performance if any, tragedy struck.

While trying to reach agreement on the sonic effects of using an ABX box, one scientist was suddenly mauled to death by hundreds of lab rats. As the rats began their horrid attack, the man was heard to yell "I'm sorry, I guess I shouldn't have suggested that we just listen to see if we hear any differences."

It is suspected that when the rats heard the scientists suggestion, they began to fear that they'd lose thier jobs and be exterminated. Said a rat, who wished to remain anonymous, "It was either him or us... and there's alot more of us."

The study has been temporarily suspended while investigators and a team of experts look into the attack.

---End of Story---

I just saw the late news, which announced. The rats have retained the top-ranking PETA lawyer. The lawyer stated: “While it is too early to have gathered all facts pertinent to this investigation, I am confident that I will be able to prove due to the Scientist’s aggressive taste in Music, which terrified the rat slated to be the first impartial observer placed into the box, my clients where pushed beyond reason.
The Rat when interviewed commented, “They just can’t understand the mechanics of my superior and delicate auditory abilities” before his Lawyer whisked him away.

And, sadly, I don't doubt that there would be people who did that and reported on an internet forum somewhere that the switchbox miraculously made their 950 sound more like a 1066.

Over and out.

PS. SLL: great post!

Maybe... and then again I'm sure that you would swoop right in to rain on their parade and tell them all what idiots they all are, since they could not possibly be really hearing any difference. Thank goodness that there are folks like you who take the time to save folks from their own perceptions.

And what... you liked SLL's post but you didn't like my post? Did you do DBT ABX on that? Are you sure your biases didn't enter in on that opinion?

Why 100%? Everyone seems to think that they can run a couple of subjects, or even themselves only, in a double-blind test and come up with some sort of answer. Yeah, sure. A positive result in a DBT speaks for itself. However, the converse, a negative result, is another creature. What kind of preliminary data and power analyses were performed to determine if the N (sample size) were appropriate?

Some extremes:

1) A person runs a DBT 100 times. He scores 53 out of 100 correct. Would anyone doubt that this is chance?

2) 1000 people run a DBT 100 times. Each of them scores between 51 and 55. If all 1000 people scored above 50%, any analysis you care to perform will tell you that the difference, though small, is real.

Now put that first person back into the second scenario. What does that score of 53 really mean? Is it still chance? If you haven't done the appropriate design, and used an appropriate N, who knows?

DBT as performed in audio shows an appalling lack of understanding what the test is actually sensitive to. It has a reasonably low level of Type I error, that is, saying a difference is present when in fact it is not. However, the probability of Type II error, saying a difference is NOT present when in fact it is, is astonishingly high...so much so that the test is useless without an N orders of magnitudes higher than those normally used in this kind of study.

A useful test must show low probability of both Type I AND Type II error. This is where the DBT/ABX crowd have their little blind spot.

If somebody can show me a DBT in audio where Type II error was even considered, much less controlled for, I'd be interested in the reference.

You're absolutely right: negative results require a much larger N to be compelling. Much more and better designed testing needs to be done.

The obvious effects that basic linear systems theory suggests a cable can have on the signal passing through it (attenuation, phase shifts, etc.) are all basic acoustic manipulations for which you can find well-established human hearing thresholds in academic journals. Measurable physical properties of well-made interconnects usually predict a signal degradation orders of magnitude below these human thresholds. (There are exceptions, though. I believe the MIT and Transparent cables both house passive filter networks which make noticable changes to the signal.)

The interconnect tests that I've read about in greatest detail were performed at DAL, using an N at which other subtle component differences were significant. He's rather chatty on USENET; I bet if you dropped him a note, he'd be happy to share more details with you.

Hm... I'm curious... if something like extreme #2 were in fact observed (highly significant p-value, but very small actual difference), how much would people be willing to shell out for interconnects?

Thanks SLL! This pretty much explains my golf game. Now I have a scientific excuse for all those triple bogeys.

This has been a very entertaining thread. I recall a similiar discussion that was equally entertaining. It was on rec.audio on the usenet.......in 1985. The debate rages on and should be resolved shortly after the cure for the common cold.

"In a double blind world the one-eyed man is king."

When I see 'then they plugged in the Wowee cables and I immediately heard a distinct improvement in the ......' that tells me the difference is so obvious, even with notoriously poor long term audio memory, that an ABX (which is a sensitive listening test) should be a snap. In theory you are correct, of course, but in practice even tiny audible differences in A and B will often cause the results to swing quickly.

In tests where noise or distortion are introduced to one source as an experiment on thresholds of audibility the swing from ~50% to ~100% typically occurs over a fairly narrow range.


Charlie

Kudos to Smart little Lena...Cleared up the
cloud thing for me...been suspecting it for years..

Everybody quit pickin on Merc..He just needs to switch to Decaf sometime before 10 A.M.

Where's the Love??

Eddyboy

or switch off of alcohol after 10PM.

brianca



[This message has been edited by brianca (edited July 03, 2002).]