JC 2 General Paper GP Tuition | GP Revision Guide

GP revision is not the final act of rereading everything you studied. It is the controlled process of deciding what still matters, retrieving it without support, finding what breaks under changed conditions, repairing the earliest useful cause and proving that the repair survives later. The best revision programme makes the learner’s system smaller, more available and more reliable.

This is the revision owner inside the SETC JC2 General Paper lane. For the everyday learning routine, use How to Study for GP. For the integrated A-Level runway, mocks, taper and final preparation, use A-Level Exam Preparation. This page owns revision itself: what to revisit, what to drop, how to schedule retrieval, how to allocate effort, how to repair and how to know when a revision target is genuinely stable.

The revision question is not “What have I covered?”

Coverage is historical. Revision needs a current-state question: What can I now retrieve, interpret and execute independently under changed conditions? A student may have “covered” climate policy in March but be unable to reconstruct the central tension in September. Another may remember every example but still misread extent questions. A third may understand Paper 2 deeply but lose accuracy late in the paper.

Revision therefore starts with state, not sequence. The order in which material was taught does not determine the order in which it should be revised.

Maren, Iona and Leonie revise different bottlenecks

Maren tends to revise by expanding. She adds examples, notes and alternative arguments. Her revision must teach pruning and retrieval. Iona tends to revisit difficult material until she feels certain. Her revision must distinguish genuine semantic weakness from perfectionistic overchecking. Leonie tends to revise by volume: more questions, more papers, more checklists. Her revision must convert results into ranked repair decisions.

The three students may receive the same score and need completely different revision plans. The score is an output; revision needs causes.

The seven-stage GP revision loop

1. Sample: gather representative evidence.
2. Diagnose: identify the first repairable cause.
3. Prioritise: rank by frequency, cost, breadth and proximity.
4. Retrieve: reconstruct knowledge or procedure without support.
5. Repair: train the replacement behaviour directly.
6. Transfer: apply it to changed material.
7. Verify: decide whether to retain, taper or retire the revision target.

Revision becomes inefficient when stages are skipped. Students often jump from sample to more sample: a weak essay is followed by another essay. Or they jump from diagnosis to rereading: “my examples are weak, so I will read more” without testing what the learner can already retrieve. The loop forces a causal sequence.

Stage 1: sample representative evidence

Use recent school work, timed practice, short diagnostic tasks and closed-note retrieval. Do not base the revision plan on one dramatic result. Look for repeated patterns across different tasks and conditions.

For Paper 1, sample question choice, interpretation, planning, argument, evidence, evaluation, language and completion. For Paper 2, sample passage understanding, inference, paraphrase, summary, connections, application, language and timing. Across both, sample retrieval, checking and recovery.

Stage 2: find the first repairable cause

The visible mistake is not always the best revision target. An irrelevant paragraph may come from a weak topic sentence; the weak topic sentence may come from a vague thesis; the vague thesis may come from a misread key term. Revision should move upstream until it reaches a cause that can be trained.

In Paper 2, an inaccurate application answer may begin with misunderstanding the passage proposition. More local examples will not repair it. Trace the chain before assigning revision time.

Stage 3: prioritise by leverage

Four variables help. Frequency: how often does the problem occur? Cost: how much does it damage performance? Breadth: does repairing it improve several tasks? Proximity: how much time remains and how quickly can the issue realistically improve?

A repeated scope error can damage an entire essay and therefore outrank a rare grammar slip. Weak claim-strength tracking may affect inference, paraphrase, summary and application, giving it broad leverage. Near the examination, a stable but slow skill may be more realistic to optimise than a very large knowledge gap.

Stage 4: retrieve before review

Before opening notes, reconstruct. Revision should reveal what is available without the support that will be absent in the examination. Write the issue tension, mechanisms, examples and limitations from memory. Recreate the Paper 1 question-choice routine. Explain the Paper 2 semantic checks.

Then review only what the retrieval attempt shows is weak. This changes rereading from a default behaviour into a targeted repair.

Stage 5: repair the mechanism directly

If the weakness is question interpretation, revise questions. If it is thesis calibration, revise theses. If it is evidence integration, revise paragraphs. If it is summary selection, revise target boundaries and point selection. The training unit should match the weak decision.

Full papers remain necessary for integration, but they are often inefficient for initial repair because many other variables hide the mechanism.

Stage 6: transfer to changed material

A correction on the original question can be supported by memory of the feedback. Change the topic, wording, passage or context. Delay the retest. Reduce prompts. If the learner still performs the replacement behaviour, the repair is becoming transferable.

Maren’s stronger thesis on a rewritten essay is not enough; she needs to produce a calibrated thesis on an unseen domain. Iona’s corrected paraphrase is not enough; she needs semantic fidelity on a different passage. Leonie’s new pacing plan is not enough; it must survive a full timed segment.

Stage 7: verify and retire

Revision targets should leave the active queue. A weakness that remains stable across changed tasks, delay and reduced support can move to maintenance. If it returns under one specific condition, revise that condition rather than reopening the entire topic.

Retiring targets is essential. Without retirement, revision becomes an ever-expanding list of everything the student has ever done wrong.

Build a revision dashboard with three active priorities

Each priority should contain five fields: evidence, trigger, first repairable cause, replacement behaviour and next verification. Keep only three priorities active unless there is a compelling reason to add another.

Example: “Paper 1 question choice — evidence: two recent essays misread comparison burden — trigger: familiar topic word — cause: content recall before wording analysis — replacement: command/terms/scope/burden audit before shortlist — verification: three unseen sets plus one timed essay.”

Revision is resource allocation

Time spent on one target is time not spent elsewhere. Revision should therefore compare marginal value. Ten more hours on an already strong issue area may add less than two hours repairing Paper 2 summary distortion. A third full mock may add less than three short transfer tests.

This is why equal-time revision is rarely optimal. Allocate according to current evidence, not a desire to make the timetable look symmetrical.

Separate knowledge revision from performance revision

Knowledge revision restores concepts, mechanisms, examples and current evidence. Performance revision restores the ability to interpret, choose, write, read, compare, compress, apply and finish. A learner can be strong in one and weak in the other.

If an essay improves immediately after the teacher reminds the student of examples, knowledge access is implicated. If examples were available but the essay still drifted, performance revision deserves more attention.

Separate Paper 1 revision from Paper 1 content accumulation

Paper 1 revision does not mean collecting more topics. It means ensuring that topic knowledge can become an argument responsive to unseen wording. The sequence is question → interpretation → thesis → roles → evidence → evaluation → conclusion.

Revise the earliest unstable link. More evidence cannot compensate for a thesis that answers a different question.

Paper 1 Revision Gate 1: question interpretation

Use sets of prompts without writing full essays. For each, identify command, key terms, scope, relationship and burden of judgement. Write one tempting misinterpretation beside the correct reading. This makes the danger visible.

Verification requires changed domains. A learner who reads technology prompts accurately but misreads abstract culture questions still has a conditional interpretation weakness.

Paper 1 Revision Gate 2: question choice

Rank questions by finishability rather than familiarity. Compare interpretation confidence, thesis clarity, distinct roles, evidence depth, evaluation potential and time risk.

Revision should include rejected-question reasoning. Explaining why a question is risky strengthens decision rules and prevents emotional choice.

Paper 1 Revision Gate 3: thesis calibration

Write three thesis strengths for the same prompt and match each to the evidence required. Introduce a serious counterexample. Revise the strongest thesis that survives honestly.

A thesis that becomes “it depends” needs the decisive conditions named. A thesis that collapses under one obvious exception needs narrower scope.

Paper 1 Revision Gate 4: paragraph roles

Plan functions before examples. Label each paragraph in six words or fewer. If two labels are essentially the same, the essay may contain duplication.

Revision can use old scripts: erase the examples and reconstruct the role sequence. If the argument disappears, evidence may have been carrying structure that should have existed independently.

Paper 1 Revision Gate 5: evidence retrieval

Retrieve examples by mechanism, not only topic. Give the student “institutional trust”, “externality”, “distribution” or “innovation” and ask for cases from different domains. This increases portability.

Remove examples that can be named but not explained. Revision should improve the active bank, not simply reactivate every historical note.

Paper 1 Revision Gate 6: evidence integration

Take old paragraphs and highlight factual detail separately from inference. If the paragraph is mostly fact, compress. Require a sentence stating what the example permits the writer to claim and a sentence identifying the relevant limit.

Then substitute a different example. The analysis should adjust where context changes rather than remain generic.

Paper 1 Revision Gate 7: evaluation

Revise evaluation by criteria: effectiveness, fairness, legitimacy, feasibility, scale, distribution, reversibility, time horizon and evidence quality. Choose one that materially changes the paragraph.

Delete decorative caveats. If the “however” does not alter the judgement, it is not doing enough intellectual work.

Paper 1 Revision Gate 8: conclusion

Write conclusions from plans before writing full essays. State decisive criterion, strongest remaining limitation and final hierarchy. This tests whether the argument actually reaches a judgement.

When old conclusions merely repeat introductions, revise the body’s evaluative sequence as well; the weak ending may be a symptom of an essay that never compared competing considerations.

Paper 1 Revision Gate 9: timing

Time each stage separately. Slow completion can come from question choice, planning, example retrieval, sentence restarting or overdevelopment. Revise the slow stage rather than simply forcing faster writing.

Use checkpoints and a triage rule. When behind, remove lower-value factual detail before removing reasoning or conclusion.

Paper 1 Revision Gate 10: language

Build a personal error bank from actual scripts. Prioritise recurrent high-cost patterns: sentence boundaries, vague reference, imprecise modality, awkward collocation, repeated words or tangled clauses.

Rewrite real sentences and retest later. General grammar review should not displace the specific patterns actually suppressing clarity.

Paper 2 revision starts with meaning control

Paper 2 questions differ, but many failures share an upstream source: the learner has not preserved the proposition accurately. Revision should therefore begin with source voice, claim strength, logical relationship and textual evidence.

Use the dedicated Paper 2 owner for detailed mechanics. At revision level, the question is which meaning-control stage currently breaks and how to restore it.

Paper 2 Revision Gate 1: passage mapping

Revise central claim, movement, source voice and qualification. Use short passages and require a compact map. If maps become essays, reduce detail. If they are only topic labels, increase proposition precision.

Transfer across unfamiliar domains so the map is structural rather than topic-dependent.

Paper 2 Revision Gate 2: referent tracking

Use dense sentences containing pronouns, demonstratives and abstract references. Replace them with full referents before answering. This repairs meaning chains that can corrupt later inference.

Retest under time once accuracy stabilises. The learner should not need to rewrite every referent in the real paper; the slow drill builds the internal habit.

Paper 2 Revision Gate 3: source voice

Distinguish what the author endorses, reports, concedes, questions, rejects or qualifies. Students who flatten source voices can produce plausible but incorrect answers.

Revision can use stance labels on short extracts before integrating the skill into full responses.

Paper 2 Revision Gate 4: paraphrase

Identify proposition structure before changing wording. Preserve quantity, certainty, causality, comparison and condition. Use contrastive pairs: one accurate paraphrase and one subtly distorted paraphrase.

Ask the learner to explain exactly which semantic property changed. This creates a reusable checking rule.

Paper 2 Revision Gate 5: inference

Build inference gradients: stated, strongly implied, possible and unsupported. Require the textual trigger. Students with rich background knowledge must learn not to substitute what is generally plausible for what this passage licenses.

Revision is complete when claim strength matches evidence strength reliably under changed topics.

Paper 2 Revision Gate 6: summary target

Rewrite the target in plain language before hunting points. Define inclusion and exclusion boundaries. Revise target recognition independently if irrelevant points are a recurring problem.

Do not spend the whole revision session on paraphrase when selection is the first failure.

Paper 2 Revision Gate 7: summary point selection

Distinguish main reason, explanation and illustration. Cluster genuine overlap before compression. Use short extracts so the student can complete many selection decisions.

Later, reintroduce time and full-passage complexity.

Paper 2 Revision Gate 8: summary compression

If selection is strong but expression slow, give preselected points and train concise reconstruction. Preserve logical distinctions. Speed should increase only after semantic accuracy is stable.

Retest in integrated summary because compression under discovery is harder than compression of provided points.

Paper 2 Revision Gate 9: connections

Compare propositions along one dimension: cause, responsibility, severity, solution, value or consequence. State the relationship before selecting textual evidence.

Revision should remove vague “both are similar” language and replace it with exact relational claims.

Paper 2 Revision Gate 10: application

Use source proposition → mechanism → contextual feature → evidence → judgement of fit. Build context knowledge around institutions, demographics, economy, geography, regulation, media and trust rather than isolated local facts.

Retest by applying the same passage idea to contrasting contexts. If the answer changes sensibly, the learner understands the mechanism.

Paper 2 Revision Gate 11: timing

Measure where time disappears: passage orientation, evidence location, answer wording, summary, connections or application. A learner may be “slow at Paper 2” for very different reasons.

Use closure rules for low-value polishing and protect later tasks. Revision should reproduce the paper sequence periodically because timing interactions matter.

Paper 2 Revision Gate 12: checking

Build a personal review order. Common high-value checks include claim strength, passage attribution, summary omission, unearned inference and application without contextual evidence.

Use a change-justification rule so checking does not destabilise correct answers.

Revise issue knowledge as a network

Do not revise topic folders in isolation. Use concepts to connect them. Trust can link public health, media, technology and government. Externalities can link environment, transport, finance and public health. Legitimacy can link law, institutions, protest and regulation.

Network revision creates more retrieval routes and prepares the learner for prompts that combine domains unexpectedly.

Use the issue-revision ladder

1. Recall the issue title only.
2. State the central tension.
3. Reconstruct mechanisms.
4. Retrieve examples.
5. Retrieve limitations.
6. Generate a counter-position.
7. Answer an unseen question.
8. Compare the issue with another domain.

The ladder lets the learner see exactly where knowledge stops. If step three fails, do not waste time testing step seven until the model is repaired.

Use source-free revision first

Begin with blank paper. Only after recall should the learner consult notes, articles or model answers. Source-free revision exposes availability honestly.

Then use sources surgically. Fill missing mechanism, update a volatile fact, resolve a disputed interpretation or strengthen a weak contrast case.

Use reverse revision from questions

Instead of choosing a topic and reviewing all notes, choose an unseen question and ask what knowledge it demands. Retrieve the relevant concepts and examples from memory, then inspect gaps.

This method tests whether knowledge is organised for use rather than for storage.

Use concept prompts for cross-topic retrieval

Write concepts on cards or a random list: trust, agency, incentive, inequality, scale, rights, legitimacy, uncertainty, resilience, innovation. Draw one and retrieve examples from several domains.

Maren gains flexibility; Iona gains comparative precision; Leonie gains faster retrieval under an unfamiliar surface topic.

Revise current examples with a freshness gate

1. Is the factual core still accurate?
2. Has later evidence changed the interpretation?
3. Is the mechanism still relevant?
4. Is there a better current or comparative case?
5. Does this example still add a distinct function?

Do not replace examples merely because they are older. Replace when explanatory value, accuracy or contextual relevance has materially weakened.

Revise volatile knowledge more often than stable knowledge

Current policy status, emerging technology capabilities, ongoing conflicts and economic conditions can change quickly. Concepts, historical mechanisms and established institutional structures usually change more slowly. Store the two layers differently.

Frequent verification belongs to the volatile layer. Do not waste revision time repeatedly checking stable material that is already accurate and retrievable.

Revision should compress the active example bank

Near major examinations, the active bank should become smaller. Keep examples that are accurate, well understood, distinct and portable. Move weaker cases into archive.

This reduces retrieval competition. Maren’s ten examples about one mechanism can become three comparative cases she genuinely owns.

Use comparison pairs

Pair cases where one important condition differs. Similar policy, different institutional capacity. Similar technology, different regulatory environment. Similar economic shock, different safety net. Similar cultural product, different audience.

Comparison pairs generate evaluation and make revision more analytical than memorising isolated cases.

Use revision spacing deliberately

Return to weak material sooner and stable material later. The interval itself is information. If a concept disappears after three days, the model may be weak. If it remains available after weeks and transfers successfully, maintenance can be lighter.

Spacing should create effort, not total loss. Adjust from evidence rather than following one universal calendar.

Use interleaving after repair

Blocked revision is useful while the learner is rebuilding a skill. Once stable, mix it with other tasks. Interleaving forces diagnosis: which procedure applies here?

A mixed revision set might combine one unfamiliar essay prompt, two paraphrases, one summary target, a concept-retrieval task and a short evaluative paragraph.

Use difficulty ladders

Increase challenge gradually. Question interpretation can move from concrete wording to abstract concepts and nested comparisons. Paraphrase can move from simple claims to concessions, causal chains and modal qualification.

Difficulty ladders show where reliability breaks. They prevent one hard item from being misread as global failure.

Use support ladders

Begin with teacher prompts if necessary, then checklist, then self-prompt, then independent task. Record the minimum support needed for success.

A strong revised answer produced only with detailed prompting is not yet examination-ready.

Use time ladders

Repair accurately first, then shorten the allowed time. If accuracy collapses immediately, the procedure may not yet be automatised enough for pressure.

Time ladders are safer than imposing full exam speed from the first correction attempt.

The three-dimensional verification gate

• Surface change: different topic, wording or passage.
• Time change: delayed retest.
• Support change: fewer prompts.

Require success across all three before retiring a major revision target. This provides stronger evidence of durable learning than immediate correction.

Revision and mocks have different jobs

Revision changes the system. Mocks measure the integrated system. A mock can reveal a weakness, but repeating the mock format is not automatically the repair.

After a mock, stop and enter the seven-stage revision loop. Only simulate again after the important findings have produced changed behaviour.

Use mocks as hypothesis tests

Before the paper, state what you are testing: question-choice stability, late-paper accuracy, summary timing, unfamiliar-topic transfer, checking or recovery. Afterward, review that hypothesis before the total score.

This makes simulation information-rich and prevents the revision programme from becoming an endless sequence of scores.

Revision after school prelims

Prelims provide integrated evidence under meaningful pressure. Begin by separating repeated weaknesses from one-off events. Compare with earlier scripts. Identify which problems were knowledge, decision, execution or recovery failures.

Do not automatically rebuild the system. Preserve what held. Change the smallest set of high-leverage elements that explains the result.

Revision after a good prelim result

A good result can also mislead. Ask whether the question happened to match favourite topics, whether Paper 2 passages were especially comfortable and whether completion remained stable. Retest key processes under changed conditions.

Success is evidence, not immunity from diagnosis.

Revision after a bad prelim result

Do not turn one result into an identity statement. Decompose it. Which component fell? Which process failed? Did the weakness repeat? Was the paper unusually difficult for the learner’s profile? What is the first repairable cause?

Then rank the response. High-frequency, high-cost, broad-transfer weaknesses first. Rare anomalies can be monitored without displacing the plan.

Build revision around the performance floor

Best-day performance shows the ceiling. Revision should care deeply about difficult-day performance. Which procedures disappear when the topic is unfamiliar, the student is tired or the passage is abstract?

Raise the floor by practising those destabilising conditions deliberately after the basic skill is secure.

Track leading revision indicators

Useful indicators include retrieval latency, proportion of accurate mechanism sentences, question-choice time, number of duplicated paragraph roles, semantic paraphrase errors, summary target errors, completion rate, tutor prompts and successful delayed retests.

Marks remain important, but these indicators reveal movement before the total changes dramatically.

The weekly revision allocation meeting

Once a week, spend ten minutes deciding what changes. Which target is now stable? Which repeated error deserves promotion? Which knowledge area needs freshness review? Which mock finding has not yet been repaired? Which task should be removed?

Revision plans should be responsive. A fixed twelve-week timetable that ignores evidence is administration, not diagnosis.

The monthly revision prune

Retire stable errors, redundant examples, duplicated resources and rituals that no longer improve learning. Consolidate issue maps. Check volatile facts. Compare Paper 1 and Paper 2 allocation.

The active revision system should become cleaner as the examination approaches.

A six-week revision architecture

Week 1 — Baseline and triage. Sample both papers, issue retrieval and timing. Choose three active risks.

Week 2 — Knowledge access. Repair issue maps, example mechanisms and spaced retrieval. Maintain both paper skills.

Week 3 — Paper 1 leverage. Repair the highest-cost Paper 1 decision and verify on changed topics.

Week 4 — Paper 2 leverage. Repair the highest-cost meaning-control stage and verify under time.

Week 5 — Integration. Interleave, use realistic timed segments and one purpose-led mock where useful.

Week 6 — Compress. Retire stable targets, retrieve the active bank, simplify cues and move into the next preparation phase.

A two-week emergency revision architecture

When time is short, do not attempt complete coverage. Day 1 diagnoses. Days 2–4 repair the most expensive Paper 1 and knowledge-access weaknesses. Days 5–7 repair the most expensive Paper 2 weakness. Day 8 uses an integrated test. Days 9–11 repair findings. Days 12–13 retrieve and compress. Day 14 protects readiness rather than creating novelty.

The exact sequence can change, but the principle remains: diagnose, repair, transfer, integrate, taper. Short timelines make prioritisation more important, not less.

The Revision Laboratory: one hundred failure states and repairs

1. Rereads notes and feels prepared. Close notes first. Free recall reveals availability. Reread only gaps.

2. Revises topics in teaching order. Reorder by current weakness and transfer value.

3. Revises only favourite issues. Use coverage and retrieval evidence to expose blind spots.

4. Revises every issue equally. Allocate by frequency, cost, breadth and proximity.

5. Adds examples when retrieval is weak. Reduce input and restore access first.

6. Memorises examples without mechanisms. Require cause-process-outcome explanation.

7. Knows mechanisms but no evidence. Add one concrete case per major function.

8. Keeps stale examples. Run freshness gate on volatile evidence.

9. Replaces useful old examples just for age. Judge explanatory value, not novelty.

10. Has too many similar examples. Retain comparative or functionally distinct cases.

11. Revises by highlighting. Convert to retrieval questions and reconstruction.

12. Revises by copying notes. Delay note-making until after memory-based explanation.

13. Can recite issue map only in fixed order. Use generative recall with changed cues.

14. Cannot retrieve across topics. Use concept prompts and comparison pairs.

15. Cannot retrieve under time. Build accuracy first, then use time ladders.

16. Paper 1 revision means full essays only. Compress tasks to the weak decision.

17. Paper 1 revision avoids full essays entirely. Add periodic integration tests.

18. Question choice is instinctive. Rank using finishability dimensions.

19. Familiar topic overrides wording. Force wording audit before retrieval.

20. Unfamiliar topic is rejected immediately. Translate into concepts before deciding.

21. Thesis is vague. Name criterion and conditions.

22. Thesis is absolute. Stress-test with strongest counterexample.

23. Plan has examples but no roles. Plan functions first.

24. Plan has roles but weak evidence. Reconsider question or narrow claim.

25. Paragraphs duplicate. Label roles and merge repetitions.

26. Paragraphs are unrelated. Build dependency and criterion.

27. Evidence becomes narrative. Cap factual detail and require inference.

28. Evidence is generic. Add one accurate concrete anchor.

29. Evaluation is decorative. Use a criterion that changes judgement.

30. Counterargument is weak. Steelman before responding.

31. Conclusion repeats thesis. State earned hierarchy and limiting condition.

32. Essay unfinished. Trace stage timing and protect conclusion reserve.

33. Essay finishes early but shallow. Allocate more time to analysis, not more speed.

34. Grammar revision is generic. Use personal error data.

35. Vocabulary revision is word lists. Revise function, collocation and precision.

36. Paper 2 revision means full papers only. Isolate meaning decisions.

37. Paper 2 revision means drills only. Reintegrate under realistic load.

38. Passage maps are too detailed. Reduce to claim, movement and qualification.

39. Passage maps are topic labels. Increase proposition precision.

40. Referents are missed. Use slow substitution drills, then time ladder.

41. Author and reported voices collapse. Label stance explicitly.

42. Paraphrase uses synonyms mechanically. Reconstruct proposition first.

43. Paraphrase copies too much wording. Explain in plain language before writing.

44. Paraphrase changes certainty. Audit modality.

45. Paraphrase changes causality. Audit relationship.

46. Inference imports background knowledge. Require textual trigger.

47. Inference is too cautious. Compare evidence strength with answer strength.

48. Summary target is unclear. Rewrite inclusion/exclusion boundaries.

49. Summary selects illustrations. Separate main reason from example.

50. Summary duplicates points. Cluster before compression.

51. Summary merges distinct conditions. Preserve logical difference before shortening.

52. Summary is accurate but slow. Train expression on preselected points.

53. Connections compare keywords. Compare propositions along one dimension.

54. Connections exaggerate disagreement. Identify exact dimension of difference.

55. Application repeats passage. Add contextual feature and evidence.

56. Application dumps local facts. Anchor every fact to source mechanism.

57. Application forces fit. Allow reasoned non-transfer.

58. Paper 2 starts slowly. Reduce first-pass annotation.

59. Paper 2 deteriorates late. Protect later time and practise fatigue conditions.

60. Checking changes correct answers. Require specific justification for changes.

61. No checking time remains. Find earlier time leaks and protect reserve.

62. Revision plan has ten priorities. Rank and keep three active.

63. Same priority remains for months. Examine whether the repair task is wrong or verification never occurs.

64. Priorities change every day. Require repeated evidence before reallocation.

65. Student studies what feels urgent. Compare urgency with leverage.

66. Student revises what feels comfortable. Put active risk first.

67. Student revises only weak skills. Protect minimum maintenance for stable components.

68. Student does a mock every week regardless of repair. Space simulations around learning cycles.

69. Student avoids mocks entirely. Use partial simulation, then integrated testing.

70. Mock score is treated as diagnosis. Decompose into process families.

71. Good mock causes complacency. Retest under changed conditions.

72. Bad mock causes panic. Compare with repeated evidence and preserve stable procedures.

73. Prelim produces complete system rebuild. Change smallest evidence-supported component.

74. Prelim errors are never revisited. Convert into active risks and delayed retests.

75. Revision sessions have no stopping condition. Define target state before starting.

76. Revision sessions are too short to integrate. Add periodic longer blocks while keeping short repair work.

77. Revision sessions are always long. Use micro-retrieval and maintenance tasks to sustain frequency.

78. Student revises late into sleep loss. Protect cognitive quality as part of revision.

79. Student collects predicted topics. Convert predictions into optional unseen prompts, not curriculum.

80. Student learns a new essay template near exam. Use novelty gate; retain proven architecture unless evidence demands change.

81. Student changes all examples near exam. Refresh only those whose accuracy or usefulness materially weakened.

82. Student refuses all new information near exam. Maintain light updates on genuinely consequential developments.

83. Student cannot state active risks. Revision remains externally managed; rebuild dashboard ownership.

84. Student can state risks but chooses favourite tasks. Put leverage before preference.

85. Student understands feedback but cannot reproduce it. Add constrained production and faded support.

86. Immediate retest succeeds. Delay and change surface before retiring target.

87. Delayed retest fails. Reopen mechanism; do not simply repeat correction.

88. Changed-task retest fails. Skill may be cue-bound; increase transfer variation.

89. Skill works only with checklist. Use support ladder to fade prompts.

90. Skill works independently but slowly. Use time ladder.

91. Skill works quickly but inaccurately. Restore accuracy constraint before speed.

92. Student has no evidence of improvement. Track leading indicators, not only marks.

93. Leading indicators improve but marks are flat. Continue transfer while checking whether scoring bottleneck shifted.

94. Marks improve but process indicators do not. Check for favourable question conditions before declaring stability.

95. Student’s best score rises but floor stays low. Train variance and recovery conditions.

96. Floor rises but ceiling plateaus. Add stretch work after stability.

97. Final week still contains ten active repairs. Compress to highest-value cues and maintenance.

98. Final week becomes passive rereading. Keep retrieval and representative output alive.

99. Final day becomes panic acquisition. Protect known system and sleep.

100. Examination looks unfamiliar. Use the revised operating system: interpret, retrieve, choose, execute, recover.

Forty revision drills

Drill 1: Blank-page issue reconstruction.

Drill 2: Mechanism-only example recall.

Drill 3: Limitation recall.

Drill 4: Concept-to-three-domains retrieval.

Drill 5: Comparison-pair explanation.

Drill 6: Volatile-fact freshness check.

Drill 7: Unseen-question deconstruction.

Drill 8: Finishability ranking.

Drill 9: Three-strength thesis calibration.

Drill 10: Role-only plan.

Drill 11: Paragraph dependency test.

Drill 12: Example substitution.

Drill 13: Evidence-to-inference highlighting.

Drill 14: Steelman counterargument.

Drill 15: Criterion comparison.

Drill 16: Conclusion hierarchy.

Drill 17: Personal-language rewrite.

Drill 18: Passage movement map.

Drill 19: Referent substitution.

Drill 20: Stance labelling.

Drill 21: Proposition paraphrase.

Drill 22: Modality audit.

Drill 23: Inference gradient.

Drill 24: Summary target gate.

Drill 25: Point clustering.

Drill 26: Preselected-point compression.

Drill 27: Cross-passage dimension comparison.

Drill 28: Context-mechanism application.

Drill 29: Counterfactual application.

Drill 30: Stage-by-stage time trace.

Drill 31: Closure rule under a short timer.

Drill 32: Bounded skip-return practice.

Drill 33: Error-family clustering.

Drill 34: First-cause tracing.

Drill 35: Delayed retest.

Drill 36: Changed-surface retest.

Drill 37: Reduced-support retest.

Drill 38: Mixed-topic interleaving.

Drill 39: Purpose-led partial mock.

Drill 40: One-page operating-card reconstruction.

The revision audit: eighty questions

• Can the learner state three active revision priorities?
• Is each priority supported by repeated evidence?
• Is the first repairable cause known?
• Is the replacement behaviour concrete?
• Is a verification task scheduled?
• Are priorities ranked by leverage?
• Are stable targets retired?
• Does retrieval precede rereading?
• Are knowledge and performance revision distinguished?
• Are Paper 1 and Paper 2 allocated separately?
• Can question interpretation be tested without full essays?
• Can finishability be explained?
• Are theses calibrated?
• Are paragraph roles distinct?
• Can examples be retrieved by mechanism?
• Do examples have limits?
• Can evidence be substituted?
• Does evaluation change judgement?
• Can conclusions state hierarchy?
• Are Paper 1 time leaks located?
• Are language errors personalised?
• Can passage maps remain compact?
• Are referents tracked?
• Are source voices preserved?
• Does paraphrase preserve quantity?
• Does paraphrase preserve modality?
• Does paraphrase preserve causality?
• Are inferences evidence-led?
• Is summary target defined first?
• Are summary points distinguished from illustrations?
• Is overlap clustered?
• Can compression preserve conditions?
• Are connections proposition-based?
• Is application mechanism-based?
• Are Paper 2 time leaks located?
• Is checking risk-based?
• Can old issues be reconstructed from blank paper?
• Can concepts retrieve examples across domains?
• Are volatile facts dated?
• Is the example bank being pruned?
• Are comparison pairs available?
• Does spacing reflect stability?
• Is interleaving introduced after repair?
• Are difficulty ladders used?
• Are support ladders used?
• Are time ladders used?
• Do targets pass changed-surface verification?
• Do targets pass delayed verification?
• Do targets pass reduced-support verification?
• Are mocks separated from repair work?
• Does each mock have a hypothesis?
• Are mock findings converted into priorities?
• Are prelim findings compared with earlier evidence?
• Is a good prelim retested?
• Is a bad prelim decomposed?
• Is performance floor tracked?
• Are leading indicators tracked?
• Does weekly review reallocate time?
• Does monthly review remove work?
• Can a six-week plan be adapted?
• Can a two-week plan prioritise ruthlessly?
• Is practice volume constrained by repair capacity?
• Are resource changes evidence-led?
• Are predictions treated as optional prompts?
• Is late-stage novelty filtered?
• Is current knowledge maintained lightly near exams?
• Is sleep protected?
• Can the learner diagnose independently?
• Can the learner choose the next revision task?
• Can the learner explain why a target is retired?
• Can the learner identify a conditional weakness?
• Can the learner distinguish ceiling from floor?
• Can the learner recover after one difficult item?
• Can revision cues be compressed near the exam?
• Does the final plan preserve both papers?
• Does the system get smaller as reliability rises?
• Can the learner use the operating system on an unfamiliar paper?
• Is revision producing independence rather than dependence?
• Is revision changing performance rather than merely consuming time?

Where to go next

Return to the JC2 General Paper hub. Use How to Study for GP for the everyday study system and A-Level Exam Preparation for the broader final-year runway. When revision identifies a specialist Paper 1 or Paper 2 mechanism as the first weak link, route to that canonical owner rather than duplicating it here.