How to Study for Exams: A Plan That Beats Cramming

Start four to six weeks out, split your subjects across short repeated sessions, mix them rather than blocking one per day, and keep the final week for practice papers only. Cramming fails because it loads everything into a window too narrow to survive one bad day.

What does a study plan that survives actually look like?

A plan that survives is one you can lose a day to and still follow. That means fewer topics per session than you think you can manage, and slack built in on purpose.

Most study plans fail the same way. They allocate every available hour, so the first missed evening puts you behind, and being behind is what makes people abandon the plan and cram instead. Plan for about seventy percent of your available time and the plan absorbs a bad week instead of collapsing under it.

Three things belong in it and nothing else: which topics exist, when each one gets touched, and which days are empty on purpose. Anything more detailed becomes admin you will stop doing by week two.

How many weeks before the exam should you start?

Four to six weeks for a normal exam period, counted backwards from the first paper rather than the last. That gives every topic three or four passes with real gaps between them, which is where the durability comes from.

Work backwards, not forwards. List the topics, count the weeks, divide, and then protect the final week: it is for practice papers and weak spots, never for meeting material you have not seen before. If new content is still arriving in the last week, the plan started too late and the honest move is to triage rather than pretend.

The reason earlier beats longer is that the gaps do the work, not the hours. Reviewing a topic four times across a month beats four hours on it in one sitting, and the spacing schedule covers exactly how far apart those passes should sit.

How do you split time between subjects?

Not evenly. Split by the gap between where you are and where you need to be, which is rarely the same as how much you like the subject.

Rank every topic into three buckets after an honest test, not after a feeling: solid, shaky, absent. Absent topics get the most sessions, shaky ones get the most frequent, and solid ones get short confirmation passes so they do not quietly decay while you are busy elsewhere.

The trap is spending your best hours on the subject you enjoy. Enjoyment tracks competence, so your favorite subject is usually the one that needs you least. Give your sharpest time to the bucket that scares you, and use the tired end of the day for confirmation passes on what you already know.

How do you handle several exams in the same week?

Sequence them backwards from the timetable, not by subject preference. The subject you sit first stops getting new sessions earliest, and the one you sit last keeps a slot in every week including the exam week itself.

The mistake is treating an exam week as a single deadline. If you have papers on Monday and Friday, they are two different deadlines four days apart, and the Friday subject needs sessions on Tuesday and Wednesday while everyone around you is celebrating being finished.

Two practical rules. Never schedule a heavy new topic in a week where you sit a paper, because the paper will take the session and the topic will not get another. And keep the last session before each paper light: a confirmation pass, not a rescue attempt.

How do you know the plan is working?

By what you can produce cold, a few days after a session, not by how many hours you logged or how much of the plan you ticked off.

Set a checkpoint at the end of each week: take three topics you covered earlier that week, close everything, and write what you remember. If two of the three come back with effort, the plan is working. If they come back instantly, your gaps are too short and you are spending sessions on what you already know. If they do not come back at all, the sessions were too passive rather than too few.

Hours studied is the metric everyone tracks and the one that tells you least. It measures input. The checkpoint measures output, which is the thing the exam will also measure.

What is interleaving and how do you do it without chaos?

Interleaving means switching between related topics inside one session instead of spending the whole session on one. It works because the switching forces you to choose the right method each time, which is what an exam actually asks of you.

This is the part almost no revision advice gets right, and it is the single change most likely to move your grade.

Rohrer, Dedrick and Stershic (2015) gave seventh grade students math worksheets containing either a low or a high dose of interleaved problems. The heavier dose produced higher scores on tests given after delays of one day and of one month. Same problems, same total practice, different order.

What interleaving looks like in one session

Take three topics from the same subject that use different methods and rotate them in blocks of twenty to thirty minutes. Not three separate subjects, which just fragments your attention. Three topics that could plausibly appear on the same paper.

The rotation matters more than the length. What you want is to arrive at a problem without already knowing which method it needs, because that is the decision an exam makes you take and a blocked session never trains it.

Why it feels worse while it works better

Interleaved practice feels harder and less productive. Blocked practice feels smooth, because after four problems of the same type your brain stops choosing a method and starts executing one.

That smoothness is the illusion. It measures how warmed up you are, not how much you will retain. If a session felt effortless, that is information about the session, not about your knowledge. Expect interleaving to feel like going backwards for the first week.

When should you start past papers?

Earlier than feels comfortable, and not as a final rehearsal. Start once you have covered roughly half the material, using papers to find gaps rather than to measure yourself.

There are two distinct uses and mixing them wastes both. Open book, untimed, with notes: this is a diagnostic, and it tells you which topics to promote to the shaky bucket. Closed book, timed, one sitting: this is rehearsal, and it belongs in the final two weeks.

Doing only the second kind is a common and expensive mistake. You get a score, you feel bad, and you learn almost nothing about what to fix. Mark your own work against the scheme rather than reading the answers, because the marking is where you find out that you knew the content and lost the marks on presentation.

What should you do in the last 48 hours?

Nothing new. The final two days are for retrieval on material you have already met, light practice, and sleep.

Learning a new topic at this point costs more than it returns: it is the least consolidated thing in your memory and it will be the first to go. The exception is a topic that is genuinely absent and certain to appear, in which case you learn the minimum viable version and accept it will be shallow.

Practically: one timed paper the day before, not the morning of. A pass over your shaky bucket using active recall rather than rereading. And a hard stop in the evening, because the alternative undoes the day.

What if you are already behind with one week left?

Triage, and do it in writing. With one week, you cannot cover everything, so the decision is which topics you abandon, and making that decision deliberately beats making it by accident at two in the morning.

Rank topics by likelihood of appearing multiplied by how far you are from knowing them. Take the top half and ignore the rest, genuinely ignore, not "if there is time". Then run the compressed version of everything above: spread your passes across the days you have, retrieve rather than reread, and keep sleeping.

One week of spread practice still beats one weekend of cramming, even though both are late. Two sessions of three hours on different days beat one block of six, and that holds all the way down to a five day horizon.

Does sleep matter more than one more hour?

In the last stretch, usually yes. Sleep is when the day of study gets consolidated, so trading it for revision hours can cost you more than the hours return.

Sleep is not passive downtime for memory. Rasch and Born (2013), reviewing several decades of work, describe sleep as actively supporting the consolidation of newly encoded memories rather than merely protecting them from interference.

The practical version is unglamorous. Keep a consistent sleep window across the whole revision period rather than fixing it the night before, since one good night after three bad ones does not undo the deficit. Take real breaks between sessions, away from the desk. And treat the all nighter as what it is: borrowing from the exam to pay for the evening.

Where to start this weekend

Write down every topic, mark each one solid, shaky or absent after actually testing yourself on it, and count the weeks you have left. That is an hour of work and it is the hour that determines whether the rest of the plan is real.

Then book the first three sessions in a calendar, mixed rather than blocked, and leave one day of the week empty on purpose. If you are rebuilding how you study more broadly, a simple method that works puts this plan back alongside the rest.