On Thu, 12 Jan 2006 19:11:07 +0000, Signal wrote:

"Forwarder" emitted :

Stewart Pinkerton wrote:

On Wed, 11 Jan 2006 13:42:08 +0200, Forwarder wrote:


Don Pearce wrote:


OK, I've had a bit of a think - and I've written a protocol, which I
believe would be a basis of fair testing of audio components to
resolve such issues as cable sound etc.

I invite all here to read what I suggest, and let me know if I have
either missed something, or am being unfair on one direction or
another.

http://www.donepearce.plus.com/odds/dbt/


How many weeks did you spend working on this?

"The proctor should be on the skeptical side of the argument."

Is that so they can influence the participant? This is "fair"?

No - it is so that they will not collude with the participant. This is
necessary. The proctor's job is to sit and watch that the participant
perform his tasks fairly.

"..the person who makes the cable changes [..] should also be
skeptical"

As above. Shouldn't these people be neutral?

No. That is why there is an observer who is on the believer side.
Think about this.

"..subject and the proctor leave the room"...

Leave and enter the room every trial? Won't that cause disruption? How
will this and the associated delay affect participants mood and
memory?

Memory of what? He will listen to the sound and decide whether it is
sparkly or not - or whatever he has identified. He could clearly do
this over the time gap necessitated by the cable change when he was
sighted, so no problem. Why would you want the participant to stay in
the room while the cable was changed anyway - that would be really
stupid.



Two points :

1. What you describe may be a challenge, a duel, a bet... it is
certainly not a scientifically thorough investigation of anything.

It is as close as I can get to scientific in the domestic environment
which prompted the assertion of difference. If the test were removed
to a lab, the participant could reasonably claim that the changed
environment adversely affected his judgment. Don't let the best be an
enemy of the good.

2. Get rid of that ABX snake oil ******** - never proven to work under
these sort of conditions - use a verified DBT protocol instead.

NOt sure what you mean - please explain.

PS Make sure you have a statistically valid number of participants.
Let's start at 200 and work upwards from there.

You misunderstand the test and its purpose. I have said before several
times, the purpose is not to see if cables have an audible effect. It
is to see if a person who claims to hear a difference does in fact
hear it, or merely imagines he heard it.

One person, not 200 - unless you can find 200 cable sound hearers, fit
them in a living room and prevent them from cribbing off each other
while they make their identifications. And while you're at it, have
them do a thousand trials instead of twenty for better statistical
validity.

Any other bright ideas?

d

Pearce Consulting
http://www.pearce.uk.com