Effective revision is not a particular colour of highlighter or a fixed number of hours. It is a repeatable cycle: choose a specific result, try to retrieve or apply it, check the attempt, repair the cause of errors, return later and use new evidence to plan again.
The method changes by subject. The cycle does not.
Begin with the course, not a random resource
Before choosing a technique, identify:
- the current specification, syllabus or teacher-defined course;
- the knowledge and task types you are expected to perform;
- trustworthy materials for checking;
- recent evidence of what is secure and what is not.
For GCSEs and A levels, confirm the awarding organisation and current specification. A useful video for one board can still omit or frame material differently from another. Official specifications and assessment materials establish the course; teachers and reliable resources help you learn it.
Create a simple topic map:
| Topic or skill | Current evidence | Next test | Check source |
|---|---|---|---|
| Quadratic graphs | Can plot points; unsure about turning points | Sketch and explain two fresh graphs | Current course material and checked solutions |
| Source comparison | Selects details but links are vague | Five-minute comparison plan | Teacher feedback and board criteria |
| Cell division | Recalls stages but cannot explain significance | Blank diagram plus “why” questions | Specification and textbook |
Do not colour every topic red, amber or green from feeling alone. Add the evidence and date that produced the judgement.
The six-step revision cycle
| Step | Your question | Useful output |
|---|---|---|
| 1. Plan | What exactly will I be able to do by the end? | A bounded task and trusted check source |
| 2. Attempt | What can I produce without looking at the answer? | Recall, explanation, calculation, plan or written response |
| 3. Check | What is correct, incomplete or wrong? | Marked evidence, not a feeling |
| 4. Repair | What caused the most useful error? | One explanation, example or focused practice set |
| 5. Return | When will I try a fresh item? | A later review placed in the week |
| 6. Review | What changed and what remains? | Keep, adjust or replace the next task |
EEF guidance describes learning as a cycle of planning, monitoring and evaluating. That is a useful framework, but it does not tell every learner to use the same technique. AQA also tells learners to use their specification, past papers and teacher guidance; details must come from the awarding organisation for your actual course.
Step 1: plan a result, not a subject
“Revise chemistry” is too large. Better session outcomes include:
- label a required diagram from memory, check it and repair missing labels;
- solve six questions on one maths method and classify each error;
- plan two defensible arguments for a Literature question;
- recall ten key terms, then use the missed ones in fresh examples;
- complete one timed section and analyse where time was lost.
Name the source you will use to check: an official mark scheme, specification, teacher feedback, worked solution or reliable course material. Without a check, fluent-looking notes can hide a gap.
Test the size of the task
A session outcome is probably too broad if you cannot answer all three:
- What will I produce?
- How will I check it?
- What will I save when I stop?
Split “revise the causes of the First World War” into something such as “retrieve four factors, explain two links between them, check against the course material, then write the weakest link as tomorrow's prompt”.
Step 2: make an unaided attempt
Close the notes for the part you are testing. Retrieval might be a blank-page explanation, flashcard answer or diagram. Application might be a new calculation, source analysis or essay plan. The Open University describes active learning as doing something meaningful with information rather than only receiving it.
For a careful testing protocol, use active recall without checking too early. If the skill is completely new, learn and study a worked example first; testing cannot replace initial teaching.
Choose an attempt that resembles the learning demand:
| Course demand | Appropriate attempt | Incomplete substitute |
|---|---|---|
| Recall accurate knowledge | Blank-page retrieval or short-answer prompts | Rereading highlighted notes |
| Select a mathematical method | Mixed fresh problems | Repeating the example with numbers changed visibly |
| Analyse a source or text | Question-led annotation and explanation | Listing techniques without an interpretation |
| Construct an argument | Plan, paragraph or full response | Memorising one model answer |
| Perform under time | Timed sample after knowledge and method are ready | Timing every early practice attempt |
Step 3: check accurately
Use a simple code:
secure: correct and explainable without help;partial: correct direction, but missing knowledge, precision or a step;repair: wrong, blank or dependent on help;check: your source is unclear, so ask a teacher or tutor rather than inventing certainty.
For essay subjects, self-marking has limits. Compare with official criteria and examples, then seek feedback on a specific feature. For maths and science, show your method so that checking can locate the error.
Check in a different colour or separate column so the first attempt remains visible. If you replace the original answer, you erase the evidence needed to diagnose it.
Step 4: repair the cause
Do not copy the whole answer and call the topic complete. Classify the cause:
- missing knowledge;
- misunderstood concept or method;
- misread command or information;
- incomplete explanation or imprecise language;
- accuracy or checking error;
- time or task-selection problem.
Choose a repair that matches. A missing definition needs retrieval; a method gap needs a worked example followed by a new problem; a weak argument needs planning and feedback.
A complete worked revision cycle
Suppose a GCSE Maths learner repeatedly loses marks on simultaneous equations.
- Plan: solve four mixed questions and identify the first unsupported step in each error.
- Attempt: complete all working without a model open.
- Check: use checked solutions; two are correct, one has a sign error and one uses elimination incorrectly.
- Repair: study one elimination example, explain why the multiplier is chosen, then solve one new question.
- Return: schedule three mixed questions two days later without the example.
- Review: if the method transfers but sign errors remain, separate accuracy checking into the next goal.
The session does not end with “simultaneous equations revised”. It ends with a narrower, evidence-based decision.
Steps 5 and 6: return, then make a decision
Schedule a fresh attempt after some forgetting can occur. There is no universal perfect interval. The spaced-practice schedule shows how to move the next review closer when recall is weak and further away when it is secure.
At the weekly review, ask:
- Which tasks produced trustworthy evidence?
- Which error repeated across questions?
- What became more independent?
- What requires an explanation rather than more solo practice?
- What is the next smallest useful test?
Use the balanced study week planner to fit the cycle across subjects. Revision is working when your next decision becomes more precise—not when every available hour is filled.
Common revision traps and their replacements
| Trap | Why it hides learning | Replace it with |
|---|---|---|
| Rewriting all notes | Produces a complete-looking product without testing use | Retrieve a structure, then add only missed or corrected material |
| Making hundreds of cards | Splits ideas into fragments and creates an unmanageable queue | A small set covering important knowledge plus application prompts |
| Completing papers without review | Measures performance but does not repair causes | Error classification, focused repair and comparable retry |
| Watching explanation after explanation | Can create familiarity without independent performance | Pause, predict, attempt and explain before continuing |
| Revising only favourite subjects | Feels fluent but ignores consequential gaps | Allocate priorities from current evidence and deadlines |
| Measuring only hours | Rewards presence rather than learning | Record output, support used, check and next return |
Copyable revision-session record
Course/topic:
Result for this session:
Attempt made without help:
Check source:
Secure evidence:
Most useful error and likely cause:
Repair completed:
Fresh retry:
Next return and reason:
Question for teacher/tutor, if needed:
Keep the record brief. Its purpose is to improve the next decision, not create another set of revision notes.
Sources
- Education Endowment Foundation — Metacognition and self-regulated learning (2026-07-23)
- The Open University — Active learning (2026-07-23)
- AQA — Revision guidance (2026-07-23)



