Reply #106: certainly it can't

[OnTheOtherHand]

I thought Seger's "LOL" was pretty much on target.

ETA: I've encountered people who think they can rule out non-response bias in exit polls because they once interviewed people, or once taught research methods, etc. -- so I'm inclined to suppose that tpu is not so much exaggerating his experience as its relevance. (Of course, for all I know, he might be understating it. Hard to tell in the absence of substantive arguments.)