Sleep science: cycles, clocks and the evidence
Sleep architecture, the two-process control of sleep (sleep pressure and the circadian clock), what sleep is for, and how to read survey statistics critically. With a correlational investigation of sleep and reaction time.
Read
Sleep architecture: non-REM sleep has three stages, and stage 3 is deep sleep; REM sleep follows. The sequence stage 1 → 2 → 3 → REM repeats several times a night. Early in the night deep-sleep periods are long and REM periods short; later, REM lengthens and deep sleep shortens. The first REM period begins about 90 minutes after falling asleep.[2][1]
Sleep pressure (the homeostatic drive): adenosine builds up while you are awake, and the drive to sleep gets stronger every hour you stay awake. That is why people sleep longer and more deeply after going without sleep. Caffeine works by blocking adenosine.[1]
The circadian clock: a roughly 24-hour biological clock is coordinated by the suprachiasmatic nucleus in the hypothalamus, which receives light information directly from the eyes. As it gets dark, the pineal gland releases melatonin, which helps make you sleepy. Evening exposure to room or device light is linked to less sleep.[1][3]
What is sleep for? Sleep appears to act as a housekeeping process that clears toxins building up in the brain during waking, and it supports forming and maintaining the pathways for learning and memory. Memory consolidation probably needs both non-REM and REM sleep; studies have linked REM deprivation with poorer recall of what was learned before sleep.[1][2]
How much? The recommendation is 8 to 10 hours per 24 hours for ages 13 to 18. In the 2015 US Youth Risk Behavior Surveys, 72.7% of high-school students (national sample) and 57.8% of middle-school students (9 states) did not get enough sleep on school nights; 2025 data put the figure at 75% of students.[3][4]
Reading the evidence: young people who sleep too little have a higher risk of obesity, type 2 diabetes, injuries, poor mental health and attention problems. These are associations: on their own they do not show that short sleep is the only cause, and survey data are self-reported, so they carry uncertainty.[3][4]
Try it: Sleep and reaction time: a correlational study
You need
- A 30 cm or 50 cm ruler
- Numbered cards (one per pupil, for anonymity)
- A record sheet
- A calculator or spreadsheet
Safety
Drop the ruler only from a small height over the catcher's hand, away from faces. Sleep data are private and anonymous; no one has to take part, and results are never used to judge anyone.
What to do
- Each pupil takes a numbered card and writes, privately, roughly how many hours they slept last night next to their number on the record sheet (optional: anyone may skip).
- Ruler-drop test in pairs: the catcher holds thumb and finger open at the 0 cm mark; the dropper releases the ruler without warning; read the catch distance in cm.
- Do 5 drops per person after one practice drop. Record the mean catch distance (shorter = faster reaction).
- Plot catch distance (y) against hours of sleep (x) for the class as a scatter graph.
- Describe any pattern. Is there a correlation? How strong?
- Evaluate: name confounding variables (practice, time of day, caffeine, attention, honesty of reported sleep) and explain why this study cannot prove cause and effect.
What happens, and why?
Sleep loss is linked to attention problems, so you might expect slower reactions after short sleep. But one night of self-reported sleep, a small class and many confounding variables make any pattern weak evidence. This is the same limitation as the national surveys: they show associations, not proof of cause.[3][4]
Play: Sleep pressure or body clock?
Each statement belongs to one of the two systems that control sleep. Is it the homeostatic sleep pressure or the circadian body clock?
Your best score is kept only on this device. Nothing is sent anywhere.
Quiz
Other ages: Ages 6–7 · Ages 10–11