My friend Chris Mooney, a freelance science journalist, has just published a fantastic article in Mother Jones on "The Science of Why We Don't Believe Science":
[A]n array of new discoveries in psychology and neuroscience has ... demonstrated how our preexisting beliefs, far more than any new facts, can skew our thoughts and even color what we consider our most dispassionate and logical conclusions. This tendency toward so-called "motivated reasoning " helps explain why we find groups so polarized over matters where the evidence is so unequivocal: climate change, vaccines, "death panels," the birthplace and religion of the president  (PDF), and much else. It would seem that expecting people to be convinced by the facts flies in the face of, you know, the facts.
Chris brings together a wide range of studies and evidence to show how our pre-existing attitudes help shape the conclusions we reach. To my mind, it's the best popular article on what we know about motivated reasoning -- check it out.
(Disclosure: I talked with Chris about the article, and my research with Jason Reifler is cited in the piece.)