On Fri, 10 Feb 2006 16:37:20 +0000 (UTC), Steven Sullivan
wrote:

Note, too, that scores tended to be higher for the Monster vs 24g (choral)
than monster vs 16 (choral) even when they didn't individually reach the
p0.05 level. With more trials some of these listeners may have been
able to score a statistically significant difference.

So, did you offer the same sort of insight after the "Zip test", when
the test had to be terminated early, right in the middle of a trend
that Zip was displaying towords showing a positive difference between
amps? Seems like the box malfunctioned right at the point where he was
becoming a "trained listener"...