What a weekly review is for
Not for studying. This is the distinction that makes it work: the hour produces no revision, no notes and no cards, and if you let it turn into a study session it stops doing its actual job. Its job is measurement and triage — finding out where you are, deciding what gets abandoned, and resetting the coming week.
Because it produces nothing visible, it's the first thing dropped when you're busy — which is precisely when it's most valuable. Protecting it requires treating it as maintenance rather than as optional planning.
The checklist
- 1
Empty the capture inbox (15 minutes)
Every photo, recording, slide deck and scribbled note from the week gets processed or explicitly deferred. Badly processed beats unprocessed — a two-line summary of a lecture you can't face is worth far more than an untouched deck you'll never return to.
- 2
Check the review queue's trajectory (2 minutes)
Not today's count — the trend. A queue growing week on week means you're adding cards faster than you're reviewing, and the fix is deleting cards, not studying harder. Left alone this is the single most common reason people abandon spaced repetition.
- 3
Score each module out of five, honestly (5 minutes)
One number per module: how confident are you that you could sit an exam on what's been covered? The number matters less than the trend across weeks, and the module that quietly slips from 3 to 2 to 1 is the one you'd otherwise notice in week 10.
- 4
Look at the calendar four weeks ahead, not one (5 minutes)
Deadlines cluster. Looking one week ahead means you discover the three-assignments week when you're in it. Four weeks of lead time is what lets you start something early, which is the only real defence against a bad cluster.
- 5
Pick one thing to abandon (5 minutes)
Explicitly. A reading you won't do, a module you'll do the minimum in, a society commitment. Unabandoned obligations don't disappear — they persist as background guilt and cost you evenings without ever being worked on.
- 6
Choose three priorities for next week (5 minutes)
Three. Not a task list — three outcomes that would make the week a success. Everything else is optional, and having decided that in advance is what stops the urgent from eating the important on Wednesday.
- 7
Do one repair (20 minutes)
One hour on the lowest-scoring module, or the backlog item that's been rolling for three weeks. One. The instinct is to fix everything, which is why most people fix nothing.
Why the trend matters more than the number
| Signal | What it means | Action |
|---|---|---|
| Review queue growing three weeks running | Too many cards, not too little discipline | Delete a third of the newest cards. This is normal maintenance. |
| One module scored 2 or below twice | It has stopped being touched, whatever you intended | Next week's repair hour goes here, no negotiation. |
| Capture inbox never empties | Processing slot isn't happening | Move the slot or shrink it. A 40-minute slot that happens beats 90 that doesn't. |
| All modules at 4, nothing to abandon | Either it's week 2, or you're not being honest | Check against an actual past paper rather than a feeling. |
| Same repair item three weeks running | It's too big to be a repair item | Break it into pieces or abandon it deliberately. |
When to do it
Friday afternoon or Sunday evening, and the choice matters more than it sounds. Friday means you finish the week knowing what next week holds and can actually rest — the review closes the loop. Sunday evening means the plan is fresh but you've spent the weekend carrying an unresolved sense of being behind.
Whichever you pick, anchor it to an event rather than a clock time: after your last Friday class, before Sunday dinner. Time-anchored habits break the first week your timetable changes, which is the standard result in habit formation.
The failure modes
- It becomes a study session. You start processing a lecture, get absorbed, and never reach the triage. Timebox each checklist item.
- It becomes planning theatre. A beautiful colour-coded plan for a week that doesn't happen. Three priorities, written in a sentence each, is the whole output.
- It only happens when things are going well. The weeks you skip it are the weeks it would have caught something. Do a ten-minute version rather than none.
- Nothing is ever abandoned. A review that only adds is an inventory, not a triage, and the list grows until the whole system feels hopeless.
- The scores are optimistic. Confidence is not calibrated — check one module against a real past paper every few weeks to recalibrate.
The version that survives a bad week
When the week has gone badly, the full hour feels impossible and skipping it is exactly wrong. Do the ten-minute version: empty the inbox as far as you can, score the modules, pick three priorities, choose one thing to abandon. Skip the repair.
That's enough to stop a bad week becoming a bad month, which is the entire function. A system's value is measured in what it does when you're not on top of things — see how to build a study system for where this hour fits in the larger structure.