Originally Posted by
Hiro11
there are so many variables and the results seem to me to fall within the margin of error for any testing. I might be wrong, but that's what I believe.
Jan appears to think that, too, so he disbelieves that any method of testing could have a smaller margin of error to any other method of testing, which is why he doubts that any other method could be superior to his. So he sticks with his. [Edited to add:] That sounded meaner than I meant it to be. If he really believes this way then what he's doing makes perfect sense. I think there
are differences among testing methods, so I try to use more robust protocols and procedures -- but maybe I'm the one who's wrong.