garym wrote: 
> That's why replications are important. It is not any single test, but
> lots of tests, as carefully done as possible, in different settings,
> same settings, different types of subjects, different time frames, etc.
> etc. With enough of this type work we can speak to not only the internal
> validity of the tests but the external validity as well.  We are rarely
> interested in the results of a single test in a single setting, even if
> such test is perfect.  And of course all our testing should be theory
> driven (if a difference is found, what is at work that might plausibly
> cause such a difference?). Otherwise we have a kitchen sink model (throw
> it all in and just test away until you find some correlation).
If the test is inherently skewed towards producing null results unless
very carefully administered by professionals with expertise in the area
of perceptual testing, then repeating it ad-nauseum will not make any
difference - you will consistently get null results


------------------------------------------------------------------------
jkeny's Profile: http://forums.slimdevices.com/member.php?userid=35192
View this thread: http://forums.slimdevices.com/showthread.php?t=103842

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

Reply via email to