Reply #60: one quibble about "methodology and polling practices"

[OnTheOtherHand]

I think it's important to keep in mind that people decide whether they are willing to participate in exit polls. It's easy to say (not that you said this) that the 2004 exit poll would have gone better if the interviewers had been better screened and better trained, but it might not have gone much better.

Election Defense Alliance tried to do a sort of gold-standard exit poll in LA County in 2008, and by their own admission, they ended up with a ferocious oversample of Republicans. Maybe too many of their interviewers looked like hippies (or whatever) -- but maybe it was completely beyond their control.

I'm sure TIA means well, but I do wish he would lay off the pseudoscience.