In article ,
Scott Dorsey wrote:
Tom McCreadie wrote:
Dave Plowman wrote:
Thing with all these tests is you can usually get the result you are
looking for - if in charge of the tests.
But the whole point - well, half the point :-) - of a
properly-conducted double blind test is that the researcher himself is
_not_ in charge of the tests; he's not allowed to be aware of the
identity pf the choices that he's presenting to the "patients", so he's
not in a position to influence the results.
You can still influence the results a lot by adjusting the time between
the samples are presented, subject selection, equipment placement, and
choice of material. ESPECIALLY choice of material.
Edison did blind demonstrations at which members of the general public
could not tell the difference between a live singer and cylinder
playback. Today's unsophisticated audiences are more sophisticated than
that, but it's a lot easier to do with a singer than a harpsichord.
Even with double blind testing, the outcome can be influenced by the
choice of material used.
As a simple example, if doing a live versus recorded test, it would be
much more difficult to reproduce an entire symphony orchestra than a
single instrument - or voice.
--
*If I agreed with you, we'd both be wrong.
Dave Plowman
London SW
To e-mail, change noise into sound.