You're right. When you're dealing with so many variables, a fool-proof double-blind control-group-equipped experiment is virtually impossible.It can be even more complicated, KarlBob. Your four conditions model assumes that there are no other major variables that affect what is being tested.