"You stick your hand in a cage, and pull out a rat," Ian Roberts, a professor of epidemiology at the London School of Hygiene and Tropical Medicine, told The Scientist. "The rats that are the most vigorous are hardest to catch, so when you pull out 10 rats, they're the sluggish ones, the tired ones, they're not the same as the ones still in the cage, and they're the control. Immediately there's a difference between the two groups."
. . .
"For ethical and cost reasons, researchers try to use as few animals as possible, which can mean minuscule sample sizes. Unblinded, unrandomized studies are the norm.
And, of course, rats aren't people. Back somewhere around 1972 or so there was an article in Scientific American about comparing animal and human studies, in which they noted that for one variety of PCB (I think it was) a rat could stand 100x the per body mass dose that a guinea pig could, and a factory accident had exposed some people to levels 1000x the guinea pig's without immediate ill effects. Googling shows different numbers, but the spread is even larger than the earlier report said: a factor of 2500 and people aren't even in the list.