Studying
Spaced repetition, explained properly
Two separate findings, usually run together, and the one that matters most is not the spacing.
Spaced repetition bundles two distinct findings, and separating them is the most useful thing you can know about it — because one is doing more work than the other, and it is not the one in the name.
The testing effect. Retrieving something from memory strengthens it far more than re-encountering it does. Trying to recall an answer and failing is more useful than reading the answer again.
The spacing effect. Reviews spread across days produce more durable memory than the same number of reviews packed together.
Both are among the better-replicated findings in learning research, across many decades and many kinds of material. What follows is what they mean in practice, including where they do not help.
The testing effect is the bigger one
If you take one thing from this page: stop rereading.
Rereading is the most popular revision method and close to the least effective. It feels productive because the material becomes familiar, and familiarity feels like knowledge. It is not — it is fluency with the page, and the exam does not have the page in it.
The alternative costs nothing and is uncomfortable, which is why people avoid it: close the book and try to say it. Write the answer out. Explain it aloud. Draw the diagram from memory. The struggle is the mechanism, not a sign it is going badly.
A useful reframing: the moment you cannot recall something is the moment the method is working. Rereading skips that moment, which is exactly why it feels nicer and does less.
Then space it out
Once you are testing rather than reading, spread the tests.
A workable schedule, without any software:
| Review | When |
|---|---|
| 1st | The same day you first learn it |
| 2nd | The next day |
| 3rd | Three days later |
| 4th | A week later |
| 5th | Two to three weeks later |
| 6th | A month later |
The intervals expand because each successful recall makes the memory more durable, so it can survive a longer gap before the next test. Anything roughly like this works — the specific numbers are far less important than the shape, and the research does not identify one optimal schedule.
The practical rule that matters more than the table: if you got it right easily, wait longer next time. If you struggled, come back sooner. That is the whole algorithm any app implements.
Where it genuinely does not help
Worth being straight about, because it is sold as universal and is not.
Understanding, as opposed to recall. Spaced repetition is a memory technique. It will reliably get a definition, a date, a formula or a vocabulary item into long-term memory. It will not make you understand why the formula is true, and drilling something you do not understand produces a memorised string you cannot use.
Understand first, then schedule. Cards made from material you have not understood are the most common way people spend forty hours and learn very little.
Essay subjects, mostly. For a history or literature paper, what you need is a small number of arguments you can deploy and evidence you can reach for. Some spaced work on dates, names and key quotations helps. Most of the value is in reading, thinking and writing, and no schedule substitutes.
Anything you will use tomorrow. If the exam is on Thursday, the spacing effect has no room to operate. Test yourself anyway — the testing effect works at any timescale — but the schedule is irrelevant.
When the cards become the task. The most common failure. Making cards is pleasant, visible, and feels like studying. Reviewing them is neither. If you have three hundred cards and have reviewed forty, the system has become the displacement activity it was meant to prevent.
Interleaving, the third effect
Usually left out, and it changes how you build a session rather than how you schedule it.
Interleaving means mixing problem types within a study session instead of doing all of one type and then all of the next. Practising A A A B B B C C C is blocking; practising A B C B C A C A B is interleaving, and across a range of studies the interleaved group tends to perform worse during practice and better on a later test.
That gap is the important part, and it is the same shape as the testing effect. Blocked practice feels much better and works less well. When every question is the same type, you stop deciding which method applies and simply execute the one you are already using — but recognising which method applies is most of what an exam actually tests.
Where it helps most:
- Maths and physics problem sets, where selecting the right approach is the difficulty
- Anything with confusable categories — similar species, similar case law, similar grammatical constructions
- Language vocabulary mixed across topics rather than grouped by theme
Where it helps least: material with only one procedure to apply, and any skill you are encountering for the first time. Learn a method blocked, then practise it interleaved.
The practical version is unglamorous: when you finish a problem set, shuffle in questions from the previous three chapters. Textbooks are organised by chapter, which quietly trains you to know the method before you read the question.
What a good card looks like
Most of the difference between a deck that works and forty wasted hours is in the writing, not the schedule.
One fact per card. A card asking for four causes of something will be half-right every time, and half-right cannot be scheduled — you will never know whether to advance it.
Ask for production, not recognition. “Which of these is the definition of X?” can be answered by elimination. “Define X” cannot.
Make the front unambiguous. “1848” is not a question. “In which year did revolutions break out across the Austrian Empire, France and the German states?” has exactly one answer, and you will still understand it in six weeks.
Include the why on the back. Not to memorise, but because the card is also your revision note. A bare answer tells you nothing when you get it wrong.
Write your own. Shared decks are tempting and much less effective, because writing a card requires deciding what matters and phrasing it — which is itself a first retrieval. A downloaded deck skips the most valuable step and encodes someone else’s understanding of the syllabus.
Building it without an app
You do not need software, and for many subjects a notebook is better because writing the answer out is itself a retrieval.
- Question on one side, answer on the other. Physical cards, a folded page, or two columns you cover with your hand.
- Three piles: today, this week, this month. Move a card up when you get it right and back to today when you do not. That is spaced repetition, complete.
- Write questions, not topics. “Causes of the 1848 revolutions” is a topic and you cannot test yourself on it. “Name three economic causes of the 1848 revolutions” is a question with an answer you either produced or did not.
Fitting it into a week
The awkward part is not the method, it is that reviews are small, frequent and easy to skip — unlike an essay, nothing goes wrong immediately when you miss one.
Two things help. Attach reviews to something that already happens daily rather than scheduling them for a time you have to remember. And keep sessions short: fifteen to twenty-five minutes of testing is more than most people manage with real attention, and our study timer will keep an honest count of what you actually did rather than what you meant to do.
If the reviews are part of a bigger deadline, the assignment planner will show you how many hours are genuinely left between now and the exam — which is usually the number that decides whether a schedule is a plan or a wish.