Why you forget what you study, and the 15-minute habit that fixes it

Think of a topic you studied carefully two months ago. Not skimmed, studied. Maybe the layers of the scalp, or the steps of the urea cycle, or the causes of a raised anion gap. Now try to write down everything you remember about it, right now, without looking.

For most people the honest answer is: about a third of it, and the part that survived is the part that was strange or memorable, not necessarily the part that matters. That feeling is not a personal failing. It is how memory works, and the people who prepare well for a huge syllabus are mostly the people who stopped fighting it and started working with it.

This article is about why we forget, why the usual way of studying is the least effective way to prevent it, and a small daily habit that changes the picture. It's long because the details are what make it usable.

The curve everyone has heard of

In 1885 the German psychologist Hermann Ebbinghaus published results from experiments in which he memorised lists of meaningless syllables and tested himself at intervals. He found that forgetting is fast at first and then slows: you lose a lot of what you learned soon after learning it, and the remainder fades more gradually. This is the forgetting curve, and although the exact numbers depend on what you are learning and how you learned it, the shape has been replicated many times since.

Two practical points come out of it. First, the most dangerous time for a new fact is the first day or two. Second, the curve is not fixed. Each time you successfully recall something, the curve for that item gets flatter. It decays more slowly the next time. Recall, in other words, isn't just a test of memory; it changes the memory.

That second point is the one that matters, and it is where most study habits go wrong.

Why rereading feels so good and works so poorly

When you reread a chapter, the words are familiar. Familiar feels like known. You nod along, you highlight, you feel a pleasant sense of coverage. But recognising a sentence when it is in front of you is much easier than producing the idea from nothing, and the exam asks you to do the second thing.

Researchers have a name for the gap: the illusion of competence. A well-known 2013 review by John Dunlosky and colleagues compared common study techniques and rated how useful they are. Highlighting and rereading, the two most popular, were rated low in utility. Practice testing (trying to recall, answering questions) and distributed practice (spreading study out over time) were the two rated high. That matches what the lab work shows and what most successful exam candidates discover on their own, often late.

The mechanism behind practice testing has its own name, the testing effect. In a much-cited 2006 study, Henry Roediger and Jeffrey Karpicke gave students a passage to study. One group reread it several times; another read it once and then practised recalling it. A few minutes later the rereaders did better. A week later, the group who had practised recall did clearly better. The act of pulling information out strengthens it more than putting it in again.

If you take a single idea from this article, take that one: retrieval is not how you check what you've learned. Retrieval is how you learn it.

Spacing: the other half

The second finding is about timing. Studying something five times in one evening is much less effective for lasting memory than studying it five times over five weeks. This is the spacing effect, and it is one of the more robust results in the psychology of learning.

Here is a way to feel why it works. When you review something right after seeing it, recall is easy and so the review does little. When you wait until you've nearly forgotten it, recall takes effort, and that effort is what strengthens the memory. The sweet spot is to review at the moment it has become a little hard, but not impossible.

That sets up a puzzle for a student with 20 subjects. You can't track, in your head, when you last reviewed several thousand individual facts. Which is exactly the job a spaced repetition system does for you.

How spaced repetition actually works

The idea is simple. Every item (a card) has a due date. When it comes up, you try to recall the answer, then rate how it went. If you got it easily, the next review is pushed far out. If you struggled, it comes back sooner. Over time, items you know drift away from your daily queue, and items you keep forgetting stay near.

NextPrep uses a scheduler in the SM-2 family, the same general approach that has been used in flashcard software for decades. In practice you rate each card with one of four answers: Again, Hard, Good or Easy. Roughly:

• Again means you couldn't recall it. The card returns within minutes in the same session and its interval resets.
• Hard means you got there with real effort. The next gap grows only a little.
• Good means you recalled it comfortably. The gap grows steadily.
• Easy means it was effortless. The gap jumps.

The cleverness is not in the maths. It is in the fact that you stop deciding what to revise. The queue decides. That removes a surprising amount of daily friction, because "what should I revise today?" is a question that quietly eats willpower.

Rate honestly or the system lies to you

The one way to break spaced repetition is to be generous with yourself. If you "kind of" knew a card and tap Good because it feels rude to tap Again, the card gets pushed out and you meet it again weeks later, having forgotten it. The scheduler can only be as good as your honesty.

A reasonable rule: if you had to think for longer than a couple of seconds, or you got the answer only partly right, it's a Hard or an Again. Nobody is grading you. The only person you can cheat here is you.

What makes a good flashcard

Most flashcard decks fail for the same reason: they ask for too much at once. A card that says "List the causes of acute pancreatitis" has no clean answer. You recall seven of nine, you're not sure whether to call that right or wrong, and you rate it randomly. The data that comes out is noise.

Good cards share a few traits.

One question, one answer. The test of a good card is that you know immediately whether you got it right. "Which enzyme is deficient in Lesch-Nyhan syndrome?" has one answer. "Tell me about Lesch-Nyhan" has no end.

Ask the thing you'd need in a question. The exam doesn't ask you to recite the chapter. It asks about mechanisms, differences and next steps. A card like "How do you distinguish B12 from folate deficiency biochemically?" is closer to a real question than "What is vitamin B12?".

Prefer contrasts. Many of the mistakes people make are between two similar things: Colles' and Smith's fractures, pemphigus and pemphigoid, Type I and Type II error. A card that makes you state the difference is worth five cards that describe each separately.

Keep a short hook. A one-line reminder on the back, such as "Antidote is N-acetylcysteine; NAPQI is the villain", can anchor the card. Not a paragraph. A single line.

Split lists. If a list matters, break it into several cards, or use a mnemonic card and test each part. A list of nine items can be nine cards, or three cards of three.

How many cards, how often

There is no magic number, but there is a principle: the daily queue should be something you can finish. If it takes more than 30 to 40 minutes, you will start skipping, and skipped days are what turn a system into a pile.

A sensible pattern is to add a limited number of new cards per day (NextPrep caps new cards per day for exactly this reason) and let the due cards come to you. When life gets in the way and you miss a day or three, don't try to clear the whole backlog in one sitting. Do the oldest and the cards you most often forget first, and let the rest trickle back in. The system is built to recover from missed days, as long as you return.

The 15-minute habit

If you do nothing else, do this: fifteen minutes of flashcard review every day, ideally at a fixed time. After a meal, before sleep, in the lift on the way to a posting. Fifteen minutes is short enough to survive a bad day and long enough to cover a few dozen cards. Over a month, that is more than seven hours of effective revision that took no willpower to schedule.

People are often surprised by how well this works, because it doesn't feel like much. That is part of the reason it works: it's cheap enough to do on the days you feel the least like it.

Where flashcards fit with everything else

Flashcards are good at one thing: keeping individual facts and distinctions alive. They are not good at teaching you something from scratch, and they are not a replacement for practising questions. Think of the pieces as a loop.

You practise questions, and the questions show you what you don't know. You read the explanation to understand why. The facts worth keeping become cards. The cards keep them alive. And the next batch of questions shows you whether it worked. Each part feeds the next, which is why we built NextPrep around that loop instead of around any one feature.

Mistakes that quietly ruin spaced repetition

Making cards you've never understood. A card can't rescue a concept you don't grasp. If you don't understand why a drug causes a cough, a card saying "ACE inhibitors cause a dry cough" is a fact with no anchor, and it will leak away. Understand first, then card the part you need.

Making far too many cards. If you card everything, the queue becomes unmanageable and you burn out. Card the things you actually forget, the high-value distinctions, the facts the exam loves. Skip what you already know cold.

Rating by how the answer felt rather than whether you got it. Feeling confident is not the same as being correct. Be strict about it.

Reviewing in the wrong state. If you are exhausted, your recall will be worse than usual and you will rate yourself down. That's fine, but don't take it as a sign that you know nothing. It's a sign you're tired.

Never looking at the cards you fail repeatedly. Some cards keep coming back. Those are the most valuable ones, because they point to something you haven't really understood. Stop and re-learn them properly, then rewrite the card in a better form.

A worked example

Take a topic from the para-clinical block: the adverse effects of ACE inhibitors.

You meet it in a question. You choose "hypokalaemia" and the answer is "dry cough". The explanation says ACE inhibitors block breakdown of bradykinin, which accumulates and irritates the airway, and that they raise potassium, not lower it.

Now you make two cards instead of one fat one. Card one: "What is the signature adverse effect of ACE inhibitors and why?" Card two: "ACE inhibitors raise or lower potassium?" Both have clean answers. Both come with a small hook ("switch to an ARB for cough"). Tomorrow they appear. If you get them right, they come back in a few days, then a week, then a few weeks. If you miss one, it returns the same day.

That is the whole mechanism. A small thing, done every day.

What to expect

The first week feels slow, because the queue is mostly new cards and you're not yet seeing the benefit. By the third or fourth week, the queue is a mix of cards at different stages and you start noticing that facts are simply there when a question needs them. That's the point at which most people stop thinking of it as a chore.

What you won't get is a guarantee. Spaced repetition makes memory more durable, but it doesn't understand things for you, and it doesn't replace thinking. It does one job well, which is to stop the things you've already learned from quietly leaving.

Where to start tomorrow

• Pick one subject and make ten good cards from questions you got wrong this week.
• Schedule fifteen minutes at a fixed time and protect it.
• Rate honestly, and let the cards you fail come back.
• Cap new cards so the queue stays finishable.
• Once a week, look at the cards you keep failing and understand them properly.

If you'd like to try it, NextPrep's flashcards use this scheduling, and the sample decks cover all 20 subjects. They are original practice material that hasn't yet been through independent expert review, and they say so. Use them to feel how the loop works, and make your own cards from the mistakes that matter to you. Those are the ones you'll remember.