Khadgar wrote:This is a superb example though. Here we have someone of conservative political bent, who upon seeing data that conflicts with what he "knows" to be true, rejects it. There's no logical reason to reject the polling data. The methodology is sound and well accepted and has been for years. He simply refuses to believe it.
Why I'm not sure. Admitting mistake isn't easy of course, but this goes beyond that I'd wager.
I haven't seen the particulars of the poll. However, you can easily get a skewed poll by localizing the search area. Also, the likelyhood of persons to take the time to answer a series of poll questions dependent on what time phone calls were made.
There's likely some validity to the poll, but it would also likely have a greater margin of error.





