Most study advice is folklore passed down from whoever got good grades at your school. Two techniques have actual research behind them, replicated across decades and populations, and they are the ones students use least: practising retrieval instead of re-consuming, and spacing that practice instead of massing it. The reason they are unpopular is also the reason they work — they feel harder than reading.
Why re-reading fools you
When you read a chapter a second time, it feels fluent and familiar, and your brain files that feeling as "known". It is not known; it is recognised. Recognition and recall are different abilities, and exams test the second one while re-reading trains the first. This is called the illusion of competence, and it is why a student can read something five times and still go blank in the hall. Highlighting is worse, not better — it is the most popular technique and among the least effective in every review that has measured it.
Active recall: how to do it properly
- Close the book and write what you know. The blank page is the whole method. Everything you cannot write down is exactly the gap you need, and it is visible in ninety seconds.
- Take notes as questions, not copies. Instead of transcribing "the mitochondria produces ATP", write "what does the mitochondria do, and why does that matter?". Notes you can read are notes you will only read.
- Answer out loud before you check. Speaking forces a complete formulation; pointing at a paragraph does not.
- Do it before you feel ready. Recalling while the material is still shaky builds far more than recalling after five passes, even though it feels much worse. The difficulty is the mechanism.
- Use past papers as the question source when you cannot write good questions yourself. A mark scheme is a free recall test with the answers attached.
Spaced repetition: the schedule
Forgetting is exponential, and each successful retrieval flattens the curve for that item. So you review at expanding intervals — the point being to meet the material roughly when you are about to lose it, not when it is still fresh. A workable hand-built schedule for a topic you learn today:
| Review | When | What it should feel like |
|---|---|---|
| 1st | Same day, a few hours later | Easy — you still have it |
| 2nd | Next day | Effortful but successful |
| 3rd | Three days later | Struggle, then recall |
| 4th | One week later | Genuinely hard |
| 5th | Two weeks later | If it sticks here, it is yours |
| 6th | Monthly until the exam | Maintenance only |
An item you get wrong goes back to the short end of the schedule. An item you get right three times in a row can be dropped — the biggest waste in spaced repetition is reviewing material you already own because it feels good to do so.
Doing it without an app
- Two-pile cards. Known and unknown. Only the unknown pile gets reviewed each day; anything you get right three separate times graduates.
- Leitner boxes. Five boxes, review frequency increasing with each box. A shoebox and a marker does this perfectly well.
- A dated list. Write each topic with its next review date and work the list top-down. This is what the app does for you, and it takes ten minutes to set up.
Where the timer fits
Retrieval is effortful in a way reading is not, and unbounded effortful work is the easiest thing in the world to abandon. So put it in blocks: 25 minutes of closed-book recall beats two hours of open-book re-reading, and the pomodoro structure is what makes the hard version tolerable. The break after each block is not a reward — it is when the consolidation actually happens.
The three traps
- Making cards instead of using them. Card creation is comfortable, productive-feeling procrastination. Cap it: make cards inside one session, then never make another one until you have used what exists.
- Card overload. Several thousand cards is a schedule you will abandon by Wednesday. Cull ruthlessly to what is examinable.
- Reviewing everything due in one sitting. When the backlog grows, triage by marks available rather than clearing the queue — a missed day is not a moral failure, it is a scheduling fact.
Neither technique needs software, and both need a bounded block and a place to record what you actually finished. That is the whole reason the Task Magnet exists here: a session attached to a named question, so at 9pm you know what you could answer cold today rather than how many hours you sat down.
