Originally Posted by potlimit

As someone with post-grad work in stats I can tell you that if the sample size is large enough, a number like %100 would make me skeptical of the whole experiment... just from a purely experimental standpoint. You'd have to expect SOME variance in SOME variable.

.