Roughly one-quarter of the kids fell into the borderline-acceptable category, meaning they reported eight hours of shuteye nightly. The overwhelming majority fell short — with 30.2 percent reporting seven hours, 22.8 percent slumbering closer to six hours, 10 percent catching a mere five hours of sleep, and 5.9 percent claiming to nod off for no more than four hours most weeknights. Just the thought makes me yawn. Certainly, schools don’t help the situation by starting classes earlier for teens than they do for younger kids — even though puberty and other developmental changes lead to adolescents needing more sleep than grade schoolers, not less.
But there could be other issues.
Like what share of teens don’t get enough sleep because they’re naturally night owls (like me) and find almost anything before 2 or 3 a.m. more interesting than slumber? http://www.sciencenews.org/view/generic/id/53220/title/Vast_majority_of_teens_are_sleep-deprived

The circadian link is to a gatewayed article; you can find a public copy at http://www.thefreelibrary.com/Sleepy+teens+haven%27t+got+circadian+rhythm.-a0134623686

In an attempt to reset the students' daily biological clocks Biological clocks, or circadian rhythms, so that they would be more alert in daytime and go to bed earlier, the researchers exposed some students in their classrooms to especially bright light between 8 a.m. and 10 a.m. Other students were exposed to muted red lighting. But the bright light neither changed students' sleep patterns nor improved their scores on tests of mood, vigor, and cognitive function cognitive function Neurology Any mental process that involves symbolic operations–eg, perception, memory, creation of imagery, and thinking; CFs encompasses awareness and capacity for judgment .

I remember. It was a very high-profile study that asserted a phase-delay in the teenage sleep cycle. The study was cited for a while in arguments to shift the school day later by a couple hours.