getting something straight in my head

One of the biggest circularities in sleep research, and the reason why it is far more popular among biologists than psychologists, is that many behvaioral studies end up saying, essentially, that sleep deprivation causes people to become sleepy. This is not revolutionary news, and people recognise that, and yet under different guises they say it again and again.

Why? The problem is this: when one studies how sleep deprivation affects cognition, behavior and performance, the fact that subjects are sleepy is mostly an irrelevant phenomenon. Performance declines for two reasons -- (1) because you're cognitively impaired, and (2) because you just want to sleep. The first is interesting because it's related to some physiological ceiling, a biochemical battery that needs to be "recharged"; the second is far more fundamental, the basal parts of the brain telling you that the first thing is happening, and that it's time to give up the ghost. Theoretically, one could be asleep, but still technically capable of functioning fine -- if one were awake.

Schizophrenia is a good analogy. Schizophrenics are often cognitively impaired - on certain tests of memory and executive function in particular -- but that impairment often goes undetected because it is completely masked by the fact that these people have lost their marbles. It's extraordinarily difficult to study brain function in very low-functioning schizophrenics, precisely because their putative cognitive deficits are convolved with the whole kit and caboodle of their other problems, including things like: "Do you even understand the instructions to the damn test?"

In the real world, of course, this doesn't always matter. If you fall asleep at the wheel, no one cares if you could potentially have been attentive had you been awake, you're still wrapped around a tree. Falling asleep trumps everything else. In other situations though -- say pulling an all-nighter before a big presentation, or flying the red-eye to a business meeting -- the two things get pulled apart: you haven't slept, but you're aroused enough that you're not likely to nod off in front of the board of directors. Those are the interesting scenarios, and those are the data that a lot of groups fail to capture and explore.