Spaced Repetition for Exams: A Student's Guide (No Apps Required)

Most students study a topic once, feel good about it, and never touch it again until the exam. Then they sit down in the hall, look at a question they "definitely covered", and nothing comes out.

That isn't a memory problem. It's a scheduling problem. You learned the thing. You just learned it at the wrong time, once, and let it rot.

Spaced repetition is the fix. It's the second most useful study technique in the research literature (right behind practice testing), it costs nothing, and you can run it with a notebook and a pen. This guide shows you exactly how.

The forgetting curve is the whole problem

In the 1880s, Hermann Ebbinghaus memorised lists of nonsense syllables, then tested himself at intervals to see how fast they disappeared. The result is the forgetting curve: memory drops off steeply in the first hours and days after learning, then flattens out at a low level.

The important part isn't the drop. It's what happens when you review.

Each time you pull the information back out of your head just as it's about to fade, the curve resets — and it resets flatter. The next drop is slower. Review again at the right moment and it's slower still. After a handful of well-timed reviews, the memory holds for weeks or months with almost no effort.

That's spaced repetition: reviewing at growing intervals so each review lands just before you'd forget.

Why cramming feels like it works (and then doesn't)

Cramming loads everything into memory in one session. On the night of the cram, you can recall most of it. That's real. The problem is that a single massed session puts you at the very top of the steepest part of the forgetting curve. Two days later most of it is gone, because you never gave your brain a second retrieval to anchor it.

Cepeda and colleagues reviewed decades of studies on this and the pattern is consistent: for the same total study time, spreading it out beats bunching it up, and the longer you need to remember something, the wider the gaps should be. Dunlosky and colleagues' 2013 review of ten common study techniques rated distributed (spaced) practice and practice testing as the two with the highest utility. Highlighting and re-reading landed at the bottom.

So if you've got an exam in six weeks, one three-hour session on a topic is close to the worst way to spend those three hours. Six 30-minute sessions spread across the six weeks is close to the best.

A concrete review schedule

Here's the guideline. It's a starting point, not a law — you'll adjust it based on how each topic behaves.

After Day 30, a review every month or so keeps it live. If your exam is inside that window, just make sure the last review lands 2–5 days before the paper.

Two rules for adjusting:

• If you nail a review easily, stretch the next gap. Day 7 felt trivial? Skip Day 14 and go to Day 21.
• If you fail a review, reset that topic to Day 1. Don't keep it on the long schedule. It hasn't earned it yet.

What "review" actually means

This is where most students blow it.

A review is not re-reading your notes. It's not watching the video again. It's not looking at the flashcard, thinking "yeah, I know that", and flipping it over.

A review is retrieval: you produce the answer from a blank state before you look at anything. Write the definition. Solve the problem. Explain the process out loud. Then check.

Spaced repetition and retrieval practice aren't two separate techniques. Spacing tells you when to study. Retrieval tells you how. Spacing your re-reading sessions is better than nothing, but spacing your retrieval sessions is where the big gains are. Roediger and Karpicke's work on the testing effect showed that a test, even an unmarked one, does more for long-term memory than an equivalent amount of restudying — and spacing multiplies that.

So every review on the schedule above should look like a mini test. For a maths topic, that's three or four problems with the worked solutions covered. For a science topic, it's writing the explanation for a process from memory. For an essay subject, it's outlining an argument for a likely question without notes.

Running it on paper

You don't need an app. Here's a system that works with one notebook and a calendar.

1. One page per topic. Write the topic name at the top, the date you learned it, and the five review dates from the schedule.
2. Under that, write 3–6 retrieval prompts. Not facts — questions. "Explain why the reaction rate doubles" rather than "reaction rate doubles because…". Leave space beneath each.
3. Put the review dates in your calendar. Phone calendar, wall planner, whatever you'll actually look at.
4. On a review day, answer the prompts on a scrap of paper, cover to check, then mark yourself honestly. Tick if you got it fully, cross if not. Two crosses out of five means that topic goes back to Day 1.
5. Keep a running list of "reset" topics. These are your real weaknesses. They get priority in your study timetable.

This takes about ten minutes a topic per review. Across a subject with twenty topics on staggered schedules, you'll be doing maybe two or three reviews a day. That's it.

Running it with an app

Flashcard apps with built-in spacing (Anki is the standard one) do the scheduling for you. Each card carries its own interval, and the app shows you only what's due today.

They're good. Two warnings:

• Cards must be retrieval prompts, not facts. A card that says "Photosynthesis" on the front and a paragraph on the back trains you to recognise the paragraph. A card that says "Write the balanced equation for photosynthesis and name where each stage happens" trains you to produce it.
• Don't let the app hide the big picture. Apps are great for atomic facts (vocab, formulas, dates) and weak for whole-problem practice. A history essay or a three-part calculus question doesn't fit on a flashcard. Pair the app with proper practice questions.

If you're preparing for a specific exam, the highest-value "cards" are exam-style questions. That's where tools like StudyAce sit: unlimited AI-generated exam-style practice papers, marked honestly (one mark for the working, one for the answer, and "not sure" scores zero), so each review is a real retrieval attempt with real feedback. Use your exam board's free past papers too — here's how to get the most out of them.

What to do with a topic you keep failing

Some topic will refuse to stick. You reset it to Day 1 three times and it still falls over at Day 7. This is normal and it's diagnostic.

A topic that won't stay in memory usually has one of three problems:

You never understood it in the first place. You memorised the steps without the reason. Go back to the source, find the one idea you're missing, and rebuild from there. Spaced repetition can't rescue a memory that was never properly formed.

Your prompts are too big. "Explain electromagnetism" is not a review prompt. Break the topic into 4–6 smaller questions and schedule those separately. Usually two of the six are the real problem and the other four are fine.

It's interfering with a similar topic. Two formulas that look alike, two historical events with similar causes. Review them side by side and write out explicitly what distinguishes them. Then space that comparison as its own prompt.

Don't just keep hammering the same card. Change something.

The mistakes that make spaced repetition useless

Reviewing recognition instead of recall. Glancing at the answer, feeling familiar, moving on. This is the big one. Familiarity feels like knowledge and isn't. Always produce before you check.

Reviewing everything equally. The topics you know well don't need Day 3 reviews. Push them out. Spend the saved time on the reset list. Spacing is about allocating effort to where the forgetting is happening.

Starting too late. The schedule needs at least two or three weeks to do anything. If your exam is in four days, this isn't the tool — this is. Start spacing at the beginning of the term, not the end.

Skipping reviews because the topic "feels done". It feels done because you reviewed it yesterday. Check again on schedule. The feeling is not the data.

Making it a project. Colour-coded systems, beautiful card decks, elaborate tracking. Ten minutes of ugly retrieval beats an hour of building the perfect system.

Starting this week

Pick one subject. List its topics. For each one, write the five review dates in your calendar starting from the day you next study it. Write three retrieval prompts per topic.

Then just follow the calendar. Don't think about it, don't optimise it, don't restart it with a better system in a fortnight. Do the reviews when they're due, mark yourself honestly, reset what fails.

In three weeks you'll notice the thing that spaced repetition does: topics you reviewed on Day 1, 3 and 7 will still be there on Day 14 without you having thought about them. That's the curve flattening. That's the whole point.