(John Atkinson) wrote in message . com...
"Michael McKelvy" wrote in message
link.net...
This is excerpted from a thread on the diyaudio.com website
I thought Arny might wish to weigh in either here or there on
the matter.

You didn't include the link to the Stereophile article on double-blind
testing. It is http://www.stereophile.com/features/141 .

John Atkinson
Editor, Stereophile

Thank you for posting this--it is very informative. I wish I had read
this a long time ago.

There are three problems here that I had suspected for a while. The
first, which isn't the main focus of the article, is generalizability,
which Clark conceded at the outset. Results are only valid for the
two pieces of equipment under test. Period.

The second is the conclusion reached about equipment from a small
number of trials. The clearest example of this was the conclusion
that PASS and Yamaha sounded the same based on a handful of trials at
Sunshine Stereo.

The third is the difficulty of identifying subtle differences in an
ABX test.

A further important fact emerged in this exchange, which it is
interesting to see has been of such longstanding duration. And that
is the tendency of this same small group of men to impugn the
objectivity and integrity of those who criticize ABX, to claim
criticisms are the result of not taking such tests, and to expound the
virtues of ABX in whatever space available no matter what the topic at
hand. This has clearly been an ongoing multimedia gesamtkunstwerk
since at least 1986. There is also a weird tendency to use
nonsensical logic and nonsequiter in response to basic logical
objections. For instance Nousaine's interpretation of the probability
of a difference being heard as a probability of a difference being
present.