You didn't answer the question, though. You didn't even postulate a testable hypothesis.

"Water freezes" is a fundamental experiment to run, yes. "Prison sentences" is complex, certainly, but it doesn't touch on ideology.

Jonah Lehrer devotes several chapters to this in How We Decide (and footnotes it, so the studies are legit, unlike that nasty Bob Dylan business). Simply put, we have evolved to focus on and pay attention to information that affirms our preconceptions, while we are evolved to tune out and ignore information that contradicts our preconceptions. We do not start as blank slates, we start as piles of hunches that, like a ball in a pachinko game, bounces off of facts on our way to the bottom… all the while avoiding things we know are going to hurt our brains. Play Rush Limbaugh to conservatives and give them a dial that adjusts the noise, they'll adjust it down. Play Rush Limbaugh to liberals and give them the same dial, they'll adjust it up.

So the better example is, say, global warming. Or its liberal doppelgänger, vaccine skepticism. Show two people with two ideologies one study and both will find things that reassure them in their biases. It's only until there is a preponderance of unavoidable, painful evidence that you will shock people out of cognitive dissonance and into rationalization. Even now, with the autism link widely discredited, vaccine denialists are focusing on other vaccines because "there haven't been enough tests yet." And even now, with the overwhelming majority of information supporting anthropocentric global warming, the skeptics are picking holes in the model about how fast it's happening as evidence that "nobody knows anything."

Sam Harris does a pretty good job of wrapping your head around it with "The Fireplace Delusion."