The 50-Second Read
Effective revision is not a technique. It is a system that repeatedly answers one question: what should this learner do next to make examination performance more likely?
That system starts with evidence, not stationery. First diagnose what the learner can and cannot currently do. Then prioritise the weaknesses that matter most. Repair missing knowledge or misconceptions. Retrieve without notes. Return after delay. Mix neighbouring methods. Apply knowledge in changed contexts. Practise examination wording. Add the clock only when it teaches something. Use past papers and mocks to expose whole-paper problems. Correct errors. Retest. Protect sleep and recovery. Replan from new evidence.
The sequence matters because students often use advanced revision tools before the foundation is ready. Full past papers cannot repair a missing algebra prerequisite efficiently. Flashcards cannot teach a student to select a method from a mixed Mathematics paper. More hours cannot compensate indefinitely for poor diagnosis.
The eduKate control question is: where does knowledge currently stop becoming marks, and what is the smallest next revision action that repairs that conversion?
One-Sentence Definition
Effective revision is an evidence-led cycle of diagnosis, prioritisation, repair, retrieval, spacing, application, examination rehearsal, feedback and recovery designed to convert existing study time into increasingly durable and independent performance.
This page is the complete-system synthesis. How Exam Revision Works owns the broad concept of turning coverage into retrieval and performance. How Revision Techniques Work owns method selection. How a Revision Timetable Works owns the scheduling logic. This guide connects those owners into one practical route from first diagnosis to examination day.
Revision Fails When It Begins With “What Should I Study?”
The better opening question is:
What is my current state?
A student who says “I need to revise Mathematics” has described a subject, not a learning problem.
A useful revision diagnosis might instead say:
- algebra signs unstable;
- chain rule understood but inner derivative omitted under time;
- reverse percentage confused with ordinary percentage decrease;
- geometry strong but decays after two weeks;
- paper completion weak in final thirty minutes.
Now revision can act.
The Complete Revision Loop
Read evidence → Diagnose → Prioritise → Repair → Retrieve → Space → Mix → Apply → Perform → Review errors → Recover → Replan → Repeat.
Each stage solves a different problem. Skipping stages can make revision look busy while leaving the true bottleneck untouched.
Stage 1: Read the Evidence
Start from real learner evidence:
- school tests;
- marked homework;
- prelims;
- practice papers;
- teacher comments;
- self-tests;
- timed sections;
- oral explanations;
- error logs.
Do not begin only from the syllabus checklist. A syllabus tells you what could be tested. Learner evidence tells you what currently needs work.
Stage 2: Diagnose the First Weak Link
A wrong answer can arise from many places.
- knowledge missing;
- misconception;
- retrieval failure;
- question misreading;
- method selection;
- execution;
- representation;
- timing;
- stamina;
- pressure.
Diagnostic assessment asks where the process first broke.
If a student writes:
y=(3x+1)⁵
dy/dx=5(3x+1)⁴
do not automatically prescribe fifty differentiation questions. Ask whether the student understands composition, remembers the inner derivative, recognises the structure in mixed work and preserves the step under time.
Stage 3: Prioritise
Not every weakness deserves equal time.
Prioritise using:
- dependency: does this weakness block many later topics?
- frequency: how often is it likely to matter?
- recoverability: can it improve meaningfully within the runway?
- mark impact: how many marks are being lost?
- transfer: does fixing it improve several question families?
- urgency: how near is the examination?
A foundational algebra error may deserve attention before an obscure low-frequency question type because the algebra error contaminates many topics.
Use Red, Amber and Green
- Red: missing, misunderstood or repeatedly failing—repair.
- Amber: understood but fragile, slow or cue-dependent—retrieve and practise.
- Green: accurate, independent and durable—maintain lightly.
The colours describe current learning state, not the learner’s identity.
Stage 4: Repair Before You Rehearse
If the learner does not understand the concept, more examination rehearsal may simply repeat the error.
Repair can use:
- explicit instruction;
- worked examples;
- prerequisite review;
- misconception contrasts;
- small scaffolds;
- targeted questions;
- fresh reattempts.
The aim is not to reread the chapter. It is to change the mechanism that produced the failure.
Stage 5: Retrieve Before Rereading
Once a concept has been learned, close the notes and try to produce it.
- answer from memory;
- write formula;
- explain concept;
- draw process;
- solve question;
- create example;
Retrieval practice both strengthens access and tells you what is actually available without the source.
The Retrieval-First Rule
Before you open the notes, show yourself what you can produce without them.
Then use the notes to correct what is missing.
This turns revision from repeated exposure into active evidence gathering.
Stage 6: Space the Returns
Immediate success can be misleading because material remains highly active.
Return after time:
learn → retrieve → wait → retrieve again.
Spaced practice makes memory prove that it can survive the gap between lesson and examination.
A Practical Spacing Pattern
For a newly repaired topic, an illustrative pattern might be:
- same day: accurate repair;
- two or three days later: retrieval;
- about one week later: mixed application;
- later: cumulative maintenance.
This is not a universal formula. Adjust intervals according to delayed performance. If retrieval is effortless, extend. If nearly everything is lost, shorten or relearn.
Stage 7: Interleave When Methods Are Ready to Compete
Blocked practice is useful during acquisition because the learner can stabilise one method.
But examinations do not usually announce:
This is a chain-rule question.
So later practice should mix methods.
learn the tools → distinguish the tools → choose the tool.
Interleaving trains selection, not merely execution.
Stage 8: Use Practice Questions for Different Jobs
Questions are not interchangeable.
- topic questions → acquisition and isolated repair;
- mixed questions → method selection;
- changed-context questions → transfer;
- difficulty-controlled questions → progressive load;
- exam-style questions → assessment interface;
- timed sets → speed and pacing;
- past papers → authentic integration.
See How Practice Questions Work.
Stage 9: Transfer
A learner who succeeds only when the question looks familiar is not fully exam-ready.
Change:
- numbers;
- representation;
- context;
- wording;
- problem entry point;
- near-neighbour methods.
Ask whether the same underlying schema survives.
Transfer of learning is the bridge between knowing one example and handling a new examination question.
Stage 10: Learn the Examination Interface
Subject knowledge must eventually pass through the examination form.
- command words;
- mark allocation;
- answer form;
- paper structure;
- representation;
- time;
- mark-scheme expectations.
Exam-style questions bridge subject practice into the interface the learner will actually face.
Stage 11: Add the Clock Deliberately
Timing is a separate performance variable.
Do not time everything from the start.
Use a ladder:
- untimed accurate;
- soft time target;
- single timed question;
- timed set;
- timed section;
- full paper.
Timed practice should teach pace, retrieval speed, method selection and move-on decisions—not simply create panic.
Stage 12: Use Past Papers Properly
Past papers are not just collections of difficult questions.
They provide authentic combinations of:
- question wording;
- topic distribution;
- marks;
- paper sequencing;
- method-selection uncertainty;
- time pressure;
- answer form.
Use them in a loop:
attempt → mark → diagnose → repair → reattempt related questions → return later to another paper.
See How Past Papers Work.
Stage 13: Use Mock Examinations for Whole-System Rehearsal
A mock combines many variables:
- duration;
- pressure;
- silence;
- limited support;
- paper order;
- stamina;
- uncertainty.
The value of the mock is the diagnostic information afterward.
If the learner fails late, route to stamina. If knowledge works untimed but not in the mock, inspect pressure and retrieval. If one method repeatedly fails, return to targeted repair.
Stage 14: Correct Errors as Systems
Do not merely write the correct answer beside the wrong one.
Classify:
- knowledge;
- misconception;
- retrieval;
- representation;
- selection;
- execution;
- language;
- timing;
- monitoring.
Then use error correction to change what produced the answer.
The Reattempt Rule
A correction is incomplete until the learner can perform again.
correct → explain → changed question → delayed retest.
The fresh question proves the system changed.
Stage 15: Track Progress Without Worshipping Scores
Scores matter, but they move slowly and can be noisy.
Track leading indicators:
- error recurrence;
- retrieval durability;
- prompt level;
- mixed-question accuracy;
- time-to-solve;
- questions left blank;
- confidence calibration;
- transfer.
Progress tracking should tell the plan what to do next.
Stage 16: Protect Recovery
Revision competes for finite human capacity.
More work is not automatically more learning if it produces:
- sleep loss;
- attention collapse;
- increasing careless errors;
- poor retention;
- irritability;
- stagnant performance;
Use recovery as part of the system, not as the reward after all work is complete.
Sleep Is Part of Revision
Do not routinely borrow from sleep to create extra low-quality hours.
Revision depends on attention, memory and self-control, all of which become more fragile when the learner is exhausted.
a timetable that works only by sacrificing normal sleep is not a stable timetable.
Stage 17: Replan Every Week
Revision plans should not remain unchanged for six weeks while learner state changes daily.
Weekly review:
- What moved from red to amber?
- What stayed red despite work?
- Which green topic is decaying?
- Which technique is not producing change?
- What did the latest paper reveal?
- What should receive less time next week?
- What is the new first weak link?
Revision is adaptive, not decorative scheduling.
The Revision Funnel
Early revision is broader. Late revision becomes increasingly examination-specific.
knowledge repair → retrieval → spacing → mixed practice → exam-style questions → timed sections → past papers → mocks → targeted final repairs → taper.
The learner moves from building capability to expressing capability under the final environment.
Do Not Start With Full Papers if the Floor Is Broken
A full paper is a high-integration diagnostic.
If every question exposes a different missing prerequisite, the paper can produce a long error list without efficiently repairing any of it.
Use full papers when enough of the syllabus is stable that the paper can test integration, selection, timing and stamina.
Do Not Stay in Topic Practice Forever
The opposite failure also exists.
A student can become excellent at worksheets labelled:
Chain Rule
Simultaneous Equations
Percentage
and fail the examination because the topic label was doing method selection.
Move into mixed, unlabeled and exam-style work once the component method is stable.
Notes Are a Tool, Not Revision Itself
Notes help organise and clarify. They become weak revision when most study time is spent rewriting information that is never retrieved or applied.
Use notes to:
- compress structure;
- clarify relationships;
- record common errors;
- create retrieval prompts;
- route to examples.
Then close them and perform.
Flashcards Are a Tool, Not Revision Itself
Flashcards are excellent for compact retrieval prompts when the learning target fits the format.
They are weaker for:
- long mathematical problem solving;
- extended writing;
- complex data interpretation;
- multi-step method selection;
- whole-paper timing.
Use the tool that matches the learning job.
Rereading Is Not Useless—But It Needs a Job
Rereading can be useful when the learner genuinely needs to reconstruct missing understanding or revisit an explanation.
It becomes low-value when used as the default because it feels fluent.
retrieve first → reread what is missing → retrieve again.
Effective Revision Is Not Equal Time for Every Subject
Subjects differ in:
- current attainment;
- importance to the learner’s pathway;
- exam date;
- amount of content;
- repairability;
- maintenance need.
Equal allocation can feel fair while being strategically poor.
Allocate according to need and consequence, while preventing one weak subject from consuming the entire week.
Effective Revision Is Not Maximum Hours
Hours are an input. Learning change is the output.
Two hours of distracted rereading may create less useful change than forty-five minutes of targeted retrieval and correction.
The next article asks How Many Hours Should I Revise? and explains why capacity, quality and task choice matter more than one universal number.
A 7-Day Revision Architecture
One illustrative weekly structure:
- Monday: diagnose and repair one red Mathematics weakness.
- Tuesday: English retrieval + writing practice; Science current topic.
- Wednesday: return to Monday Mathematics after delay; mixed questions.
- Thursday: Science retrieval and changed-context application.
- Friday: lighter green-topic maintenance and error-log review.
- Saturday: timed section or past-paper block plus diagnosis.
- Sunday: repair from Saturday, weekly review, next-week plan and recovery.
This is an example, not a prescription. The architecture is what matters: repair, delayed return, application, authentic performance and replanning.
A Single 60-Minute Revision Block
Illustrative design:
- 5 minutes: closed-book retrieval;
- 10 minutes: diagnose and review missing part;
- 20 minutes: targeted practice;
- 15 minutes: mixed or changed questions;
- 5 minutes: mark and correct;
- 5 minutes: record what returns later.
The exact timing should adapt. A difficult essay or full paper needs a different block. The principle is to include evidence, active performance and correction rather than fill the hour with one passive activity.
The Minimum Viable Revision Day
Some days are disrupted by school, CCA, family commitments or fatigue.
Instead of abandoning revision completely, maintain one minimum action:
- ten minutes retrieval;
- one error correction;
- five flashcards;
- one Mathematics question family;
- one paragraph plan.
Then resume the normal plan without carrying every missed task forward.
The Bad-Day Rule
A bad day should not automatically create a worse tomorrow.
If work was missed:
- identify what remains genuinely important;
- delete obsolete tasks;
- move only high-priority items;
- protect sleep;
- restart normal rhythm.
A revision timetable is a control system, not a debt ledger.
Four Weeks Out: Repair and Integration
When runway remains, focus on:
- major prerequisites;
- red topics;
- retrieval;
- spacing;
- mixed practice;
- initial exam-style questions.
Full papers can be used diagnostically, but they should feed targeted repair rather than simply accumulate.
Two Weeks Out: Increase Authenticity
Shift more capacity toward:
- timed sections;
- past papers;
- paper strategy;
- stamina;
- high-frequency errors;
- green-topic maintenance.
Do not stop repairing. The balance simply moves toward performance.
Final Days: Stability, Not Panic Expansion
Near the examination, avoid opening large new learning fronts unless necessary.
Priorities:
- high-value retrieval;
- common error checks;
- light paper familiarisation;
- logistics;
- sleep;
- confidence based on real evidence.
The goal is to arrive able to express what has been built.
The Day Before
Useful activities can include:
- short retrieval of essential knowledge;
- review of known high-risk errors;
- light examples;
- packing required materials;
- confirming time and location;
- stopping heavy study early enough for normal rest.
Avoid converting anxiety into an uncontrolled late-night syllabus tour.
The Examination Morning
The system is now in performance mode.
Do not try to redesign the whole knowledge base. Use brief familiar cues if helpful, then protect attention and routine.
the last hour is not where the course is learned; it is where the learner prepares to access what has already been built.
During the Examination
Revision becomes performance.
- read the actual command;
- retrieve the relevant knowledge;
- choose the method;
- show enough working to preserve state;
- allocate time proportionately;
- move when stuck;
- check high-risk errors;
- reset after mistakes.
Performance under pressure is the final conversion layer.
Effective Revision in Mathematics
Mathematics revision should move through:
concept → worked example → accurate procedure → structural cue → mixed selection → transfer → timing → full-paper integration.
The Mathematics Learning Hub owns the subject terrain. Effective revision determines which part of that terrain needs repair and how quickly it should be returned to examination form.
Mathematics Example: Chain Rule
Diagnosis:
y=(3x+1)⁵ → student writes 5(3x+1)⁴.
Repair:
What is inside what? Outer derivative × inner derivative.
Near example:
y=(2x−5)⁻².
Delayed return: mixed differentiation two days later.
Transfer: trigonometric composite function.
Performance: timed mixed calculus set.
This is revision as a progression rather than one correction.
Effective Revision in English
English needs multiple strands:
- vocabulary retrieval and use;
- reading comprehension reasoning;
- grammar and editing;
- writing planning;
- timed composition;
- world knowledge and reading;
Do not revise English only by doing full papers. A weak inference mechanism may need targeted evidence work; a weak essay may need paragraph planning; vocabulary needs repeated encounters and retrieval.
Effective Revision in Science
Science revision should preserve both knowledge and mechanism.
- retrieve definitions;
- reconstruct diagrams;
- explain causal chains;
- interpret data;
- apply models to changed contexts;
- practise command words;
- use exam-style answers;
Knowing the word “diffusion” is different from being able to explain and apply particle movement in an unfamiliar situation.
Primary School Revision
Primary learners need shorter, clearer and more supported revision cycles.
- retrieve a small amount;
- use concrete examples;
- correct quickly;
- repeat after delay;
- keep sessions bounded;
- preserve reading, sleep and normal life.
Adults should gradually transfer planning and checking rather than expecting young children to manage a complex revision system immediately.
PSLE Revision
P5 and P6 learners can use a simplified system:
- traffic-light topics;
- one red repair per subject;
- weekly older-topic retrieval;
- PSLE-style questions;
- timed sections;
- full papers only when useful;
- error review;
- parent-supported weekly replanning.
Prelims should become diagnostic evidence, not a permanent prediction of PSLE performance.
Secondary School Revision
Secondary students should increasingly own the system.
- diagnose marked work;
- set weekly goals;
- choose techniques by problem;
- schedule retrieval;
- mix topics;
- track errors;
- use past papers;
- protect recovery.
The revision plan becomes a form of self-regulated learning.
O-Level Revision
Near O-Levels, runway is finite. Prioritise:
- high-return red/amber topics;
- exam command and answer form;
- retrieval speed;
- method selection;
- timed paper completion;
- stamina;
- common errors;
- green-topic maintenance;
- sleep and stress stability.
Do not attempt to rebuild the entire academic history of the learner in the final weeks. Repair what matters most for the performance still ahead.
The Revision Audit
- What does current evidence say?
- Where is the first weak link?
- Which weaknesses have the highest dependency or mark impact?
- What needs reteaching rather than practice?
- What should be retrieved before rereading?
- What needs a spaced return?
- Which methods need interleaving?
- What needs transfer?
- Which topics are ready for exam-style questions?
- What should now be timed?
- What do past papers reveal?
- Which errors recur?
- What should be maintained rather than intensively revised?
- Is workload protecting sleep and attention?
- What changes next week?
The Revision Traffic Light
- Red system: revision is mostly passive, generic, unmeasured or panic-driven—return to diagnosis and active performance.
- Amber system: useful techniques are present but weakly scheduled or disconnected from exam performance—add spacing, mixing, timed verification and weekly review.
- Green system: evidence determines priorities, techniques match problems, errors produce repair, support fades and authentic performance improves—maintain and taper intelligently.
The Sports Performance Crosswalk
Athletes do not prepare for competition by repeating one favourite drill indefinitely.
assess → train weakness → build capacity → integrate → rehearse competition → recover → reassess.
Revision follows the same logic. Practice evolves as the learner evolves.
The Logistics Crosswalk
A reliable operation does not allocate equal resources to every subsystem regardless of condition. It monitors, prioritises bottlenecks, repairs faults, tests the repair and preserves spare capacity for disruption.
Effective revision is academic operations management under a fixed deadline.
The Governance Crosswalk
Good governance separates goals, indicators, interventions and review. A plan is allowed to change when evidence changes.
Revision should be governed the same way: not by guilt, mood or the beauty of the original timetable, but by what the learner is now showing.
Effective Revision and AI
AI can generate quizzes, explain errors, create varied questions, summarise topics and help build schedules.
But AI can also destroy the diagnostic signal if it answers before the learner attempts.
Strong sequence:
student attempts → evidence appears → AI or human helps diagnose → targeted explanation or practice → tool closes → learner reattempts → delayed independent retest.
The learner must remain the performer.
Common Failure Mode 1: Revision Means Rereading
Familiarity rises but retrieval remains weak.
Repair: retrieve first, reread only what is missing, then retrieve again.
Failure Mode 2: Revision Means Doing Papers
Every paper reveals the same errors.
Repair: pause full papers, isolate the recurring weak link, repair, then return.
Failure Mode 3: Every Subject Gets Equal Time
High-impact weaknesses are underfunded.
Repair: allocate by priority while maintaining strong subjects cheaply.
Failure Mode 4: Notes Become the Product
Pages look beautiful but performance does not change.
Repair: compress notes, convert them into retrieval prompts and spend more time performing.
Failure Mode 5: No Delayed Return
Immediate success is mistaken for retention.
Repair: schedule spaced retrieval.
Failure Mode 6: Blocked Practice Forever
The chapter title selects the method.
Repair: interleave once component methods are stable.
Failure Mode 7: Timing Too Early
The clock hides conceptual weakness.
Repair: build accurate untimed performance first unless timing is being used diagnostically.
Failure Mode 8: Timing Too Late
The real examination becomes the first speed test.
Repair: progress through timed questions, sets, sections and papers.
Failure Mode 9: Errors Are Marked but Not Repaired
The same mistake returns.
Repair: classify, correct, use a changed question and retest later.
Failure Mode 10: Too Many Hours
Quality and sleep deteriorate.
Repair: reduce low-value volume and track output quality.
Failure Mode 11: Timetable Is Never Updated
The plan stops matching learner state.
Repair: review weekly and after meaningful assessment evidence.
Failure Mode 12: Final Week Becomes Panic Expansion
New topics, late nights and extra papers destabilise performance.
Repair: prioritise high-value retrieval, familiar performance and recovery.
What Parents Can Ask
- What does the latest evidence say?
- What is the current first weak link?
- Which topic is red, amber or green?
- What will be retrieved without notes?
- What returns later this week?
- What error is being repaired rather than merely marked?
- When does revision become timed?
- Is sleep protected?
- What changed in the plan after the latest test?
What Teachers Can Do
Help students see revision as a learning system rather than a list of techniques. Provide diagnostic information, cumulative retrieval and guidance on question types. Teach students how to interpret errors, how to move from topic practice to mixed and exam-style work, and how to use past papers as evidence. Make revision strategy increasingly learner-owned.
What Tutors Can Do
Use marked work as the entry point. Diagnose before assigning volume. Repair the first weak link. Track prompt level. Use retrieval and spacing. Branch practice according to learner state. Return quickly to authentic questions. Add time progressively. Review past papers for repeated patterns. Reduce support. Make the learner explain the plan back. The tutor should gradually become a reviewer of the learner’s system rather than its permanent operator.
Case Study 1: The 52% Mathematics Paper
A student responds to 52% by planning one full paper every night.
Diagnosis shows twelve marks lost from two recurring algebra families and eight from unfinished questions. The plan changes: three targeted repair blocks, two timed sections and one full paper at the end of the week.
The next paper improves because revision changes the bottlenecks rather than merely rehearsing them.
Case Study 2: The Chain-Rule Error
A student knows chain rule but repeatedly stops after the outer derivative. One worked example exposes the nested structure. Six targeted questions stabilise the inner factor. Two days later, mixed differentiation removes the topic cue. A week later, a timed section verifies performance.
Revision progresses through repair, retrieval, spacing, selection and performance.
Case Study 3: The Rereader
A Science learner reads notes for two hours nightly and feels familiar with every chapter. Closed-book questions reveal poor retrieval.
The new rule is retrieval first. Notes are opened only after an attempt. Study time falls slightly while delayed recall rises.
Revision becomes harder in the moment and more useful later.
Case Study 4: The Flashcard Collector
An English student has hundreds of vocabulary flashcards but rarely uses the words in writing.
Flashcard volume is reduced. Retrieval remains, but each week includes sentence production, reading encounters and timed writing. Vocabulary begins appearing spontaneously because revision now includes use, not only recognition.
Case Study 5: The Paper Grinder
An O-Level learner completes many past papers. Scores plateau. Error analysis shows the same three question families failing.
Full-paper frequency is reduced temporarily. Each error family is repaired through examples, targeted practice and changed questions. Papers then resume.
Past papers become sensors rather than punishment.
Case Study 6: The Overloaded Timetable
A student schedules every free minute and begins sleeping later. After one week, attention falls and mistakes rise.
The timetable removes low-priority note rewriting, adds buffer and protects a normal stopping time. Fewer nominal hours produce better completed work.
Case Study 7: The Strong Student With Weak Exam Conversion
A learner understands content and retrieves well but loses marks under full-paper pressure. Revision stops adding knowledge and shifts toward timing, paper order, recovery after difficult questions and stamina.
Effective revision changes when the first weak link changes.
The Complete Effective Revision Control Loop
Start from real work rather than anxiety → diagnose where learning or performance first breaks → rank weaknesses by dependency, frequency, recoverability and examination consequence → repair missing knowledge and misconceptions before rehearsing them repeatedly → close the notes and retrieve → return after enough time for memory to work → interleave once methods are individually stable → vary the surface so the underlying schema must transfer → move into exam-style questions → add time as a deliberate performance variable → use past papers and mocks to reveal whole-system limits → turn every error into a classified repair and fresh reattempt → track whether support, latency, accuracy and resilience are improving → protect enough recovery that learning remains possible → replan from the new evidence → narrow toward examination conditions until the learner can independently produce what they know on the day it counts.
Canonical Owner Boundaries
This page owns the complete practical synthesis of effective revision from diagnosis to examination day, integrating learning repair, retrieval, spacing, application, exam rehearsal, error correction, progress tracking and recovery into one adaptive student system. It connects to:
- How Exam Revision Works — the broad revision owner.
- How Revision Techniques Work — choosing methods according to the learning problem.
- How a Revision Timetable Works — the scheduling architecture.
- How Past Papers Work — authentic examination practice.
- How Progress Tracking Works — evidence used to update the plan.
Evidence and Limits
No single revision technique is universally best for every subject, learner and stage. Retrieval practice and distributed practice have strong research support across many learning contexts, while the value of interleaving, worked examples, elaboration and other methods depends on the target task, learner expertise and implementation.
Exam performance also depends on teaching quality, prior knowledge, curriculum coverage, health, sleep, assessment design and other factors. A revision system can improve the use of available preparation time but cannot guarantee a particular grade.
The strongest practical rule is diagnose before technique: do not ask which fashionable revision method is best in isolation. Ask what is currently stopping this learner’s knowledge from becoming durable, selectable, accurate examination performance—and choose the method that repairs that exact problem.
The Return Path
Return to the first question:
What should I revise?
The answer is not one chapter, one technique or one number of hours.
Revise what the evidence says is limiting you, with the method that changes that limitation, then test whether the change survives time, variation, the clock and the examination itself.
That is how to revise effectively.