By Cyril KimbiMost candidates measure their revision in hours. Hours are the one metric that does not predict a grade. Two students put in the same twelve hours a week and get different results, because what separates them is not duration but which specific things they can do under time — and that is measurable.
Why a score is almost useless on its own
Two candidates score 48%. The first missed marks evenly across every topic. The second scored well on four topics and near zero on two. Identical scores, opposite problems — and opposite remedies. The first needs broad consolidation; the second needs two chapters and could gain fifteen marks in a fortnight.
A single percentage cannot distinguish them, which is why candidates who revise by score tend to re-revise what they already know. It feels productive because the score is high in those topics, and it changes nothing.
The four numbers worth tracking
| Metric | What it reveals | What you do about it |
|---|---|---|
| Accuracy by topic | Which chapters are actually weak, rather than which feel hard | Spend the next block on the two lowest, not on everything |
| Time per question | Whether you know it or merely recognise it | Drill speed on topics you get right but slowly |
| Repeat-error rate | Whether revision is sticking at all | If errors repeat, change method — do not just redo the topic |
| Decay | Whether a fixed topic is still fixed weeks later | Re-test old topics on a schedule, not only new ones |
The second row is the one almost nobody checks. A question answered correctly in four minutes that should take ninety seconds is not a question you have mastered — it is one you reconstruct each time. In a timed paper that difference is the gap between finishing and not finishing, and it never shows up in a score.
The counter-intuitive one: slow correct answers
Worth dwelling on, because acting on it is unusually cheap.
Candidates almost always revise what they get wrong. But a paper is a fixed number of minutes, and marks lost to running out of time are indistinguishable from marks lost to not knowing. If you are consistently correct but slow on a topic, the fix is not more theory — it is repetition until the method is automatic.
That is a different activity from studying, takes far less time, and candidates skip it because the topic does not feel like a weakness.
Reading the pattern in your own mistakes
Mark a paper, then sort every lost mark into one of three causes. The proportions tell you what to do:
- I did not know this. A content gap. Study the topic, then re-test.
- I ran out of time. A speed problem. More studying makes it worse, not better.
- I answered a different question from the one asked. A reading problem — and the most common of the three among strong candidates.
A candidate whose losses are mostly the third category does not need more revision at all. They need to slow down for ten seconds per question and underline what is being asked. That is a fifteen-minute habit change worth several marks, and no amount of extra studying substitutes for it.
What the app tracks for you
Doing this by hand is possible — a notebook and honest marking works. It is just laborious enough that most people stop after two papers.
In the eStudyUniverse app, practice and assessments are recorded automatically: your results screen separates the topics you were strong on from the ones you were weak on, attempt history keeps every previous sitting so you can see whether a weak topic is improving or static, and answer review shows the worked reasoning for each question rather than just a right/wrong mark. Recommended practice uses what you have already got wrong to decide what to serve next, which is the part that is genuinely hard to replicate on paper.
The analysis is the same either way. The app removes the bookkeeping, not the thinking.
How to use this across a term
- Baseline. Sit one full paper under real conditions. Resist the urge to study first — you are measuring, not performing.
- Sort every lost mark into the three causes above.
- Pick the two weakest topics only. Breadth is the enemy here; a plan that addresses eight topics addresses none.
- Re-test those two after a week, and the whole paper after a month.
- Watch the repeat-error rate. If the same errors survive two cycles, the method is wrong — change how you are studying, not how long.
Our guide to working through past papers and marking schemes covers the marking step, which is where the data comes from in the first place.
Frequently Asked Questions
How often should I sit a full timed paper?
Roughly monthly for the full paper, with shorter timed sections weekly. More often than that and there is no time to act on the results.
Is a rising score enough to know I am improving?
Not reliably. Check whether the same errors are recurring; a score can rise while the underlying gaps stay.
What if I am correct but too slow?
Drill repetitions on that topic rather than studying it again. Speed is a separate problem from knowledge.
How many topics should I work on at once?
Two. Plans covering eight topics reliably cover none of them properly.
Can I do this without the app?
Yes — a notebook and strict marking works. The app removes the bookkeeping, not the analysis.
