magiccarpetride wrote: 
> Double blind listening tests are tricky at best because they tend to
> deliver a lot, and I mean a lot of false positives. There are well
> documented cases where sufficiently large population of double blind
> testers were reporting significant differences in the sound quality
> while in reality the experimenters were merely playing an identical
> configuration over and over and over. Funny how human psyche functions.
> Highly unreliable.

With true 'Double Blind' trials the experimenters would not know what
was being analysed (played) either - that would be done by a third and
independent party. A statistically significant number of tests would
have to be carried out and statistical analysis applied to determine the
probability that a significant result (positive or negative) had been
achieved. This is provided by using a confidence scale or in medicine a
p-value. The statistical analysis can be very complex.

Bob


------------------------------------------------------------------------
bobertuk's Profile: http://forums.slimdevices.com/member.php?userid=30376
View this thread: http://forums.slimdevices.com/showthread.php?t=94418

_______________________________________________
audiophiles mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/audiophiles

Reply via email to