Reading a page again feels like learning. It is fluent, comfortable, and almost useless. The fluency comes from recognition, not from recall: your eye passes over a sentence you have seen before and the feeling of familiarity is mistaken for the ability to produce the answer under a clock.
Unreviewed, a freshly learned fact decays quickly, with the sharpest drop in the first days after you meet it. The fix is not more reading. It is arranging to meet the fact again, from memory, at the moments when it is sliding away.
Three habits turn review into memory. Each is cheap; together they are most of the difference between a topic you have seen and a topic you own.
The model omits two things. It does not say how long to wait for any individual fact, because the right interval depends on how easily that fact came back last time. And it assumes the material was understood first; retrieval cannot rescue a rule you never grasped.
| Technique | What you actually do | Why it works |
|---|---|---|
| Retrieval | Close the page and write the rule, the formula or the answer from memory, then check | Recall effort strengthens the memory far more than re-reading, which only feels productive |
| Spacing | Return to the item after a gap, and stretch the gap each time it comes back correctly | Reviewing at the edge of forgetting is the moment that buys the most lasting retention |
| Interleaving | Shuffle topics in one set, so percentages sit beside ratios and grammar beside vocabulary | Choosing the method for each question builds the skill the HAT actually tests |
| Explaining aloud | Say the rule in plain words as if teaching a friend, and note where the words run out | The gaps in a spoken explanation are exactly the gaps in your understanding |
A simple expanding ladder covers most items. Re-attempt the item cold at each rung, and if it returns correctly, move up; if not, repeat at the same interval rather than advancing.
On a Monday, a candidate logs one error: percentage change computed on the new value instead of the original. The rule is now written on one line. Here is the review schedule that follows from the ladder.
| Rung | When | What is done |
|---|---|---|
| Day 1 | Tuesday | Redo one percentage-change question cold and write the rule from memory first |
| Day 3 | Thursday | Two fresh percentage-change questions, plus the rule stated aloud |
| Day 7 | Next Monday | One question inside a mixed set that also includes ratio and averages |
| Day 21 | Three weeks on | One percentage question in a timed mini-drill; correct without hesitation retires the item |
Four short contacts over three weeks, roughly ten minutes in total, and the trap stops recurring. The same ladder applies to a grammar rule, a vocabulary word or a geometry formula; only the item changes, not the spacing.
Which activity best strengthens a topic you have already understood?
When should each spaced review be scheduled?
You work percentages, then ratios, then percentages again inside one session. What is this called?
When should you clear the Underlayer review queue?
Without looking back: name the three review habits and the first three rungs of the spacing ladder.
The answer: retrieval, spacing and interleaving; the ladder is day 1, day 3 and day 7, then day 21.
Spaced repetition schedules each review just before you would . Retrieval practice means testing yourself rather than the material again. Mixing topics within one session is called . In Underlayer, clear the daily before starting new material.
Take one rule you keep forgetting from your error log. Write it on a single line, then schedule four cold re-attempts on the ladder: day 1, day 3, day 7 and day 21. Decide now what question you will attempt at each rung.
The rule must be small enough to fit on one line, because a line can be recalled and a paragraph cannot. "Percentage change divides by the original" is a rule; "percentages" is a topic and cannot be reviewed.
At each rung the task is the same in form: close the page, produce the rule from memory, then answer a fresh question of that type. The first rung is deliberately close, while the trace is still fragile. By day 7 the item should be embedded in a mixed set, because that is where interleaving tests whether you can still find the rule when it is not signposted.
If any rung fails, the item does not advance; it repeats at the same interval. If all four succeed, retire it and replace it with the next line on your forgetting page. Four or five such items running in parallel, refreshed weekly, is a complete personal review system, and it costs less than fifteen minutes a day.
Review is now scheduled rather than hoped for. The next lesson turns the technique into a target: how to work out the score you actually need, measure your gap to it, and project whether your current rate of improvement will close it in time.