What is the difference between active recall and spaced repetition?

Active recall is a method: you retrieve information from memory rather than re-reading it. Spaced repetition is a schedule: you revisit material at increasing intervals rather than all at once. They answer different questions — active recall is how you study, spaced repetition is when you study it again. Used together they compound; used alone, each is still better than re-reading.

Which is better, active recall or spaced repetition?

The question is malformed because they are not alternatives. If forced to pick one, active recall does more work — spaced repetition applied to passive re-reading gains you comparatively little, whereas retrieval practice on a bad schedule still produces strong learning. In practice you should do retrieval, and space it.

Do I need flashcards to do active recall?

No. Flashcards are one implementation and a poor fit for anything involving multi-step working, such as maths and physics problems. Closing the book and writing what you remember, doing past-paper questions cold, and re-attempting problems you previously got wrong are all active recall, and the last of these is the highest-value version because it targets known weaknesses instead of material you already know.

What is the best spacing interval for revision?

Roughly expanding intervals: one day, three days, a week, two weeks, a month, adjusted so that anything you get wrong resets to a short interval. The precise numbers matter far less than the principle — revisit just as recall starts to get difficult. Difficulty at the moment of retrieval is what produces the memory gain, so an interval that feels slightly too long is usually better than one that feels comfortable.

How do I use active recall and spaced repetition for maths?

Use problems rather than facts. Keep a log of questions you got wrong, re-attempt each one cold from a blank page a few days later, then re-attempt a different question on the same skill a couple of weeks after that. This is retrieval practice on a spaced schedule with the content selected by your own errors, which is strictly more efficient than reviewing everything equally.

Active recall vs spaced repetition: what they are and how to combine them

The distinction in one table

	Active recall	Spaced repetition
What it is	A method — retrieve from memory instead of re-reading	A schedule — revisit at expanding intervals
Question it answers	How should this hour be spent?	When should I see this again?
Opposite of	Re-reading, highlighting, copying notes out	Cramming everything into one block
Feels	Hard, slow, discouraging	Inefficient, because you revisit things you "already know"
Fails when	You peek at the answer before genuinely trying	The intervals are so short it becomes re-reading

They stack. Retrieval practice on a spaced schedule is the combination every study-methods summary is gesturing at, and flashcard apps are simply one way to automate both at once — which is why the two ideas got welded together in most people's heads.

Active recall: the part that does the work

Active recall means producing the information yourself, from a blank page, before you look. Closing the textbook and writing everything you remember about buffer solutions is active recall. Doing a past-paper question cold is active recall. Reading a worked solution and thinking "yes, that makes sense" is not — and the reason that distinction matters is that the second one feels considerably better.

That is the trap. Fluent reading produces a strong sense of understanding, because you are measuring recognition, and recognition is easy. The exam asks for recall from a blank page under pressure, which is a different operation entirely. Students who revise by re-reading are practising a skill the exam never tests, then blanking on material they genuinely revised.

The signature. If your revision session felt smooth and productive, be suspicious. Retrieval practice feels like failing, repeatedly, in small increments. The effort of the failed retrieval is where the memory gain comes from.

The mechanism, roughly: retrieving a memory is not a read operation. It modifies the memory, strengthening the route back to it. Struggling and then succeeding strengthens it more than succeeding easily, which is why comfortable revision is close to worthless and why a check that "feels too hard" is usually correctly calibrated.

Spaced repetition: when to come back

Forgetting is not a defect, it is the default. Everything you learn decays on a curve, and each time you successfully retrieve something, the curve flattens — you keep it longer before it fades again. Spacing exploits this: revisit just as recall becomes difficult, and each visit buys a longer interval than the last.

Practically, expanding intervals: one day, three days, a week, a fortnight, a month. Get something wrong, and it drops back to a short interval. The exact numbers are not sacred — anything that expands and resets on failure captures nearly all the benefit. What breaks it is the two failure modes at either end: intervals so short that you never actually forget (in which case you are re-reading with extra steps), and no schedule at all, which is cramming.

Cramming is worth being precise about. It works — for about a week. Massed practice produces real short-term performance and very poor retention, which is exactly the wrong trade for a two-year course examined at the end. The uncomfortable implication is that the material you feel most confident about right after a cramming session is the material most likely to be gone by November.

Where flashcards fit, and where they don't

Flashcard software (Anki and the rest) is a spaced-repetition scheduler wrapped around a retrieval prompt. For discrete facts — definitions, vocabulary, formulae, reaction conditions, anatomy — it is excellent and hard to beat.

It is a bad fit for multi-step reasoning, and this is where a lot of maths and physics students waste months. A card reading "how do you integrate by parts?" tests whether you can recite a formula, not whether you can look at an unfamiliar integral, notice it wants parts, choose u and dv sensibly, and execute without a sign error. The exam tests the second thing. Nearly everyone who "does Anki for maths" and sees no improvement has run into precisely this gap.

For working subjects the unit of retrieval should be a problem, not a fact:

Re-attempt questions you got wrong, cold, from a blank page — never by reading your corrections.
Do fresh questions on the same skill a couple of weeks later, so you are retrieving the method rather than the memory of that specific question.
Blank-page a topic: write everything you can produce about it unprompted, then check against the syllabus and note only the holes.

The efficiency question nobody asks

Both techniques tell you how and when. Neither tells you what — and choosing what to review is where most revision time is actually lost.

The default is to review everything equally: work through the syllabus, chapter by chapter, including the two thirds you already know cold. Spaced repetition mitigates this a little by pushing easy material to long intervals, but only if something is measuring which material is easy for you.

The sharpest possible answer to "what" is: the things you have already got wrong. Not the things you feel shaky about — feelings about your own competence are unreliable in both directions — but the specific questions where you actually lost marks. That list is an error log, and it is a shortlist of your weaknesses generated by evidence rather than mood.

The combination worth running. Retrieval practice (how), on an expanding schedule (when), applied to your own logged mistakes (what). Each part is ordinary on its own. Together they are close to the whole of efficient revision.

Running all three without it becoming a second job

The catch is bookkeeping. Retrieval is free, the schedule is arithmetic, but tracking what you got wrong, when you last saw it, and what is due today is real clerical work — and it is the part that collapses first in week three, right when the volume gets large enough to matter.

That is the job errorboard takes over. You photograph a question you got wrong; it reads the page, works out what went wrong, categorises it and files it by topic. The review quiz then re-surfaces each mistake on an expanding schedule and asks you to produce the answer before showing it — retrieval, spaced, on your own errors, with no scheduling to maintain. If you already live in Anki, your logged mistakes export as an Anki-ready card file so you can keep your existing deck.

For the harder half — practising a skill rather than a single question — you can generate fresh exam-style questions on any topic from your log, which solves the re-attempt problem that plain flashcards cannot: new numbers, new framing, same skill.

Summary

Active recall = produce it from memory. The method.
Spaced repetition = come back at expanding intervals, reset on failure. The schedule.
They combine; the combination is not optional-extra territory, it is the baseline.
Flashcards suit facts. For maths and physics, make the unit of retrieval a problem.
Point both at what you have actually got wrong, not at the whole syllabus.

If you change one thing after reading this, make it the last point. Method and schedule are multipliers; what you point them at decides whether you are multiplying anything worth having.