Do’s and don’ts the night before the exam

Briefly, yes — twenty to forty minutes of self-testing on your two or three weakest topics, then stop. Covering the notes and trying to produce the answer is rated high-utility in the research; rereading and highlighting are rated low. What is not defensible is a long session, because the sleep it costs has a measured price and the extra study does not have a measured benefit. And keep the whole thing in proportion: in the MIT study, sleep the night before an exam had no correlation with the mark, while sleep across the preceding month did.

Eight to ten, which is what the American Academy of Sleep Medicine advises for 13- to 18-year-olds on a regular basis. That is more than most articles on this subject recommend — six and a half to eight is the usual figure quoted, and for this age group it is simply below the guidance. The phrase that matters is “on a regular basis”: one good night rescues much less than a settled month.

No. A meta-analysis of 44 studies of one restricted night found significantly increased sleepiness and impaired sustained attention — about 29 milliseconds added to reaction time and nearly three extra attention lapses — while working memory and processing speed were statistically untouched. Read that carefully, because it is the useful part: you do not lose what you know, you lose the ability to hold attention across a long paper. In a 90-minute ZAP mathematics paper or a four-hour Matura paper, that is where the marks go.

Not if they normally drink it. In a habitual consumer most of caffeine’s apparent benefit is the reversal of overnight withdrawal rather than a genuine lift, so stopping the day before means sitting the exam in withdrawal. Keep the usual amount at the usual time and stop by mid-afternoon. What genuinely matters is the evening: 400 mg six hours before bed has been shown to cut measured sleep by more than an hour, and the people in that study did not notice.

Whatever you normally eat, at roughly the time you normally eat it. That is the whole of the evidence-based answer. The clearest effects of breakfast in the research were in undernourished children, and no study supports a specific food raising an exam mark in a well-fed Swiss teenager. Do not skip a meal, do not try something new, and do not eat a very large meal close to bedtime.

Two or three minutes of slow breathing has the best evidence of anything available acutely — a meta-analysis of 26 randomised trials found a small-to-medium reduction in stress and anxiety, with no detected publication bias — and canton Zug’s school psychology service recommends practising exactly that with your child in advance. A wakeful night is also less costly than it feels: it impairs attention rather than knowledge. What we would not do is reach for melatonin, whose measured content has been found to range from under half to over four times the label, and whose evidence base is for circadian rhythm disorders rather than nerves.

Honestly, no — not for tomorrow. The research is clear that the exam was largely decided in the preparation weeks, and we would rather say so than sell you an hour tonight. Where we do earn our place is afterwards: if tonight was frightening, or the result does not reflect the work, that is a preparation problem and it repeats itself unless something changes. Send us the paper and the last report when you are ready. Book a tutor when you want to start, or message us on WhatsApp first — there is no charge to ask and none to be assigned one. Good luck tomorrow.