This is one question that can be answered with relative certainty. First
we must decide on the type of degradation that we wish to uncover; let's say
it's odd order harmonic distortion. Start out with a signal felt to be
clean by the listener. Add distortion. Now gradually reduce the amount of
distortion until it can just barely be detected by a blinded subject using
quick switching. If a long period of evaluation is more sensitive, then it
should be even easier to detect the presence of this amount of distortion
using a long period listening technique, if that's what the subject prefers.

So, is it?

Norm Strong