cleeds,
"...Real scientists actually conduct their own tests. They don’t dismiss every listener’s casual account with a wave of the hand while muttering about the lack scientific control yet making no effort to conduct their own inquiry, to their own standards."
Not to take sides, but in some of the areas of science, it is what practically happens. "Peer" in "peer reviewed" does not have to be "the guy who did the test on exactly the same subject". They do dismiss certain study exactly based on how well it was designed. Controls, sample size, etc., matter. In fact, "methods" paragraph may be the most scrutinized one.