Alicia opens her revision plan and sees an apparently sensible arrangement. The difficult topics have the largest blocks. The topics she already understands have none. Every available session goes towards the red areas, because those are where the visible problems are.
Three weeks later, the red areas are less frightening. Yet her next practice paper does not improve as much as expected. Several marks disappear from topics she had previously handled comfortably. Her family wonders whether the new revision method has failed. Alicia wonders whether she can ever improve one part of the syllabus without losing another.
Tricia responds by giving every topic equal time. That protects coverage but leaves too little space for substantial repair. Kai Kai responds by returning to his favourite topics, where success arrives quickly. His confidence improves while the difficult work remains unfinished.
Alicia, Tricia and Kai Kai are fictional learners. Their situations and the schedules below are original teaching illustrations, not measured case studies. The problem is neither that students should stop prioritising weaknesses nor that strong topics need constant drilling. The missing idea is maintenance: a smaller, evidence-guided amount of retrieval and use that keeps an existing capability available while most attention goes elsewhere.
A revision plan must do two jobs at once. It must build what is missing and preserve what is already useful. When it counts only new gains, it can quietly lose part of its starting position.
1. The 50-second route: reduce strong-topic work rather than deleting it
A topic that has become strong usually needs less intensive work, not an automatic lifetime exemption from practice. Move it from repair into maintenance. Keep a small, spaced opportunity to retrieve and use it without notes. Increase attention only if fresh evidence shows a meaningful weakness returning.
Maintenance should be chosen by what the student must later do. A formula may need retrieval and application. A reading skill may need a fresh passage. An essay skill may need a new plan or paragraph. Reopening the notes and recognising the topic name is not the same as demonstrating that capability.
There is no universal percentage of revision time that every learner should devote to strong topics. The right balance depends on the examination runway, the number of vulnerable topics, the importance of each skill and how much normal schoolwork already rehearses it. A topic used accurately in daily lessons may need little extra maintenance at home.
Begin with one question: “When did I last use this successfully without help?” That is more informative than “Have I revised this before?” The first asks about present evidence; the second can be answered by a lesson completed months ago.
The aim is not to keep every box permanently green. It is to prevent neglect from converting secure knowledge into a new repair project while the student is busy repairing something else.
2. A strong label describes evidence, not a permanent property
Students often mark a topic strong after one successful worksheet. That label may mean several different things. The learner could solve familiar examples, explain a concept immediately after teaching, answer a mixed question independently, or retrieve the knowledge after a delay. These are not equivalent observations.
Before deciding how much maintenance is needed, record what the label actually rests on. “Correct on a fresh mixed set last week” gives stronger evidence of portable access than “understood the explanation today.” The latter is useful progress, but it may still belong in consolidation rather than maintenance.
The distinction prevents a common disappointment. A family thinks a strong topic has faded, when the topic was never tested independently beyond the teaching context. The apparent decline may be the first time cue dependence became visible.
Use a plain evidence statement instead of a colour alone: “Can solve standard equations without prompts; not yet checked after a gap.” That statement directs the next task. It also protects the learner from interpreting an amber label as a judgement about intelligence.
As evidence improves, reduce the work. If a student repeatedly handles fresh applications after delays, a full weekly chapter review may be unnecessary. Good maintenance is economical because it begins with an honest account of what is already secure. It neither celebrates mastery too early nor refuses to recognise it when it is demonstrated.
3. Why success today does not finish the memory job
A successful answer can feel like a natural stopping point. The learner has crossed from not knowing to knowing, so further practice appears wasteful. Research gives a more qualified picture.
In a vocabulary-learning experiment, Karpicke and Roediger compared what happened after items had first been recalled correctly. Continued retrieval supported later recall, whereas continued study alone did not produce the same benefit in that design. The study involved a specific task and population; it does not establish a fixed maintenance schedule for every school subject.
The practical implication is modest but important: one successful retrieval is not a certificate that future access will take care of itself. A revision plan can retain a small retrieval opportunity after success without treating the learner as if they are still starting from zero.
This is especially relevant when the examination is some distance away. A student may need the knowledge not tonight but several weeks later, after many other topics have occupied attention. The question becomes whether the skill remains available at the future performance point.
Do not translate this into endless testing. A short, useful return may be enough to reveal whether the capability is holding. If it is, move on. If it is not, inspect what has changed. Maintenance exists to support learning, not to fill every evening with repeated judgement.
4. Repair and maintenance should feel different
Repair is usually more intensive. The student may need an explanation, worked examples, guided practice, correction and repeated opportunities to build a missing relationship. Maintenance begins from a different assumption: the capability has already been demonstrated and now needs to remain accessible.
Suppose Tricia is repairing quadratic modelling. She may need a substantial session translating situations into equations and comparing possible representations. Maintaining a secure linear-equation skill might require only a few fresh equations embedded in a mixed set, followed by a brief check of any unexpected error.
Giving both topics identical treatment wastes time. It can also create the wrong emotional message. The learner feels that nothing is ever completed because every topic always returns as a full lesson.
Make the distinction visible in the plan. A repair block names the missing capability and the teaching needed. A maintenance check names the established capability and the small independent task that will sample it. The student can then see why one receives forty minutes and another receives a short return.
The duration is illustrative, not a universal prescription. Some complex skills need substantial maintenance tasks; some narrow repairs are quick. What matters is the job being done. Maintenance is not a smaller copy of every original lesson. It is a deliberately chosen opportunity to keep or inspect a capability the learner already possesses.
5. The timetable can hide losses because it records only completed work
A revision checklist usually records what was done: chapter read, worksheet completed, flashcards reviewed. It rarely records what was no longer available when needed. This creates an accounting problem.
Imagine a practice paper with two broad areas. In an illustrative first attempt, Alicia earns 34 marks from her stronger area and 12 from the weaker one. After a repair period, those become 28 and 18. The total is unchanged. The repair may have helped the weaker area while performance in the previously stronger area fell.
The arithmetic does not prove why either change occurred. The papers may differ in difficulty, coverage or marking. But it gives the family a better question than “Did revision work?” They should inspect both the gain and the loss, not only the topic receiving attention.
A strong-topic maintenance check is one way to notice such a loss before the next full paper. It creates a small observation window while the repair programme continues elsewhere.
Do not build an elaborate dashboard that consumes more time than the checking itself. A simple note can do the job: last independent use, current sample, next action. The purpose is to keep prior capability visible in the plan. Without that visibility, the loudest current weakness can occupy every session while quieter losses accumulate unnoticed.
6. Choose maintenance by future use, not by page count
A topic is a large container. The capability needed in an examination may be much more specific. Maintaining “electricity” could mean recalling definitions, interpreting circuits, using an equation, evaluating an investigation or explaining a mechanism. One easy question does not sample all of those demands.
Start with the future job. What would the student need to produce if the topic appeared tomorrow? Choose a small task that requires that production. For a calculation skill, use an unfamiliar numerical example. For interpretation, use fresh data. For a writing skill, use a changed prompt rather than rereading a previously corrected paragraph.
Avoid making maintenance so broad that every return becomes a mini-exam. Rotate the important demands over time. One check might sample retrieval; another might sample selection among methods; a later paper supplies integration. Keep the claim proportional to the sample.
A student who answers one circuit question correctly has shown success on that question, not complete immunity to every electricity problem. Equally, one mistake does not establish that the whole topic has vanished. The task should produce a useful next decision, not a grand verdict.
Page count is a poor substitute for this design. Ten pages of familiar examples can give less useful information than two well-chosen fresh questions that require the learner to recognise the structure and complete the answer without nearby support.
7. Space returns, but do not worship a calendar formula
Spacing means returning after a gap rather than placing every repetition together. Cepeda and colleagues studied how the gap between learning events interacted with the later retention interval. Their results showed that useful spacing depended on when the knowledge would be needed again. They did not provide one school timetable that fits every learner and subject.
That is why a rigid rule such as “review every topic every seven days” can be too blunt. A recently repaired fragile skill may need an earlier return. A well-established skill used elsewhere may tolerate a longer gap. An examination next week creates different priorities from one several months away.
Use a tentative return date, then adjust from evidence. If retrieval is accurate and independent across several returns, the interval can often be extended. If access becomes unreliable, shorten the interval or examine whether the skill needs more teaching rather than simply more reminders.
Do not intentionally wait for total failure because a graph of forgetting suggests that struggle is always good. The practical target is useful retrieval with feedback, not preventable frustration. Nor should every hesitation trigger immediate restudy. A brief search can be part of successful retrieval.
The calendar is a support tool. The learner’s performance remains the evidence. When the schedule and the evidence disagree, revise the schedule rather than forcing the student to fit an arbitrary sequence.
8. Maintenance is retrieval and use, not reassurance
Alicia opens her strongest topic, reads three pages and feels comfortable. That comfort may be pleasant, but the task has not tested whether she can produce the knowledge without the page.
A stronger maintenance check begins closed-book. Ask for the formula, the explanation, the first move or a short solution. Then check the response. The sequence matters because opening the answer first changes the conditions under which success occurs.
For a secure skill, the check may be brief. One accurately solved fresh problem can support continuing with the planned weak-topic repair, while a later check samples a different demand. Maintenance does not need to become a performance spectacle.
If the answer is wrong, correct it carefully. Do not repeat a mistaken rule merely because retrieval practice is generally useful. The return should contain feedback when needed, followed by another opportunity to retrieve the corrected relationship later.
There is also room for restudy. A student who discovers a genuine gap may need to revisit the explanation. The point is not that reading is forbidden. It is that the maintenance decision should begin with evidence of access rather than a feeling produced by looking at familiar material.
For families, a better question than “Did you revise your strong topic?” is “What did you try without help, and what did that show?” The answer can be short and need not include a score for every activity.
9. A missed maintenance question does not automatically reopen the chapter
Tricia makes one sign error in a topic she has handled reliably. Her first reaction is to cancel the planned lesson and start the chapter again. That may be an expensive response to a narrow slip.
Inspect the error before changing the plan. Can she explain the relevant rule? Does a fresh parallel item succeed? Did the original failure concern the concept, a transcription step or the conditions of that particular task? These questions determine whether maintenance should continue, become a short repair, or expand into broader teaching.
Use a proportionate escalation. One isolated error invites inspection. Repeated failure on the same relationship suggests a more durable weakness. Failure across several prerequisites may justify a larger rebuild. The threshold should depend on the educational cost of waiting and the importance of the skill, not on a universal number of wrong answers.
The tutor-facing Maintenance and Reopening guides provide related professional reading. For the learner, the immediate lesson is simpler: a return check should help select the size of the next action.
Do not make maintenance emotionally dangerous. If every mistake causes a whole-topic restart, students will avoid honest checking. A useful system makes uncertainty manageable by connecting it to a specific, limited response.
10. Build a maintenance floor without freezing the whole timetable
A maintenance floor is a small amount of protected attention for previously established skills. It is not a compulsory percentage and not an excuse to avoid difficult work. Its purpose is to prevent the weak-topic queue from consuming every available session by default.
Start by counting existing opportunities. School lessons, homework, mixed tuition work and ordinary reading may already rehearse a skill. Add home maintenance only where a meaningful gap remains. Otherwise the student may be asked to maintain the same capability repeatedly while another capability receives no attention.
A family might protect two short checks across a week while keeping most planned study for active repairs. Another learner might use a rotating question at the start of each subject session. A student with a heavy examination schedule may need a different arrangement. These are design options, not claims of equal effectiveness.
The protected time should have a clear task and an endpoint. “Review strong topics” is too vague. “Solve one fresh simultaneous-equation problem and explain the elimination choice” is actionable. If the response is secure, stop and return to the main repair work.
Allow exceptions for genuine urgency, but record the displaced return. A maintenance task postponed once is different from one silently deleted every week. The plan should make the next opportunity visible so that urgent work does not become permanent neglect.
11. Use a week that has both repair and preservation
Consider an illustrative learner with algebra already secure, geometry under active repair and scientific explanation recently improved. The week should not consist entirely of geometry simply because geometry has the largest gap.
On the first study day, the main block can address the geometry problem. A brief independent algebra question checks whether that earlier skill remains available. On a later day, a fresh science prompt asks for a causal explanation without the previous sentence frame. A mixed task towards the end of the week samples whether the repaired geometry can be selected without a topic label.
This is not three full subjects every evening. It is a small set of purposeful returns around the main repair programme. The exact durations should fit the student’s workload and the complexity of the tasks.
Suppose the algebra check is correct but slow. The next return may sample fluency rather than trigger a full algebra lesson. Suppose the science explanation omits the mechanism again. That skill moves into a short targeted repair. Suppose geometry still requires a prompt. It remains the main teaching priority.
The plan changes because evidence changes, not because the colours on a spreadsheet must look balanced. A timetable is useful when it carries those decisions. It becomes less useful when students spend more time rearranging it than practising the capabilities it was meant to support.
12. Protect high-dependency skills first
Some strong skills support many other tasks. Algebraic manipulation can support several mathematics topics. Reading the command accurately can affect many subjects. Retrieving common vocabulary can support comprehension and writing. These capabilities deserve attention because a small loss can spread beyond one chapter.
That does not mean they all need separate drills every day. They may already be exercised naturally in current work. The maintenance question is whether they are being used accurately and independently, not whether another worksheet has been assigned.
If a student repairing calculus repeatedly encounters algebra, the teacher can inspect the algebra within those solutions. This supplies maintenance evidence without adding a separate session. If the algebra is being completed by the tutor, however, the exposure does not demonstrate that the student can still carry it alone.
A useful planning question is: “What would become harder if this skill weakened?” The answer helps rank maintenance when time is limited. A narrow fact used rarely may need a different return schedule from a frequently used operation that sits inside many later problems.
Do not turn dependency into an excuse to practise only foundations forever. Advanced tasks still need their own teaching and transfer work. The aim is to preserve the supports while extending the structure, so that new learning does not rest on capabilities the timetable has stopped observing.
13. Mathematics maintenance should include method selection
A topic can remain fluent in isolation while becoming harder to recognise in a mixed paper. Maintaining only execution therefore leaves a gap.
For example, Kai Kai can solve a pair of simultaneous equations quickly. His maintenance check should occasionally require him to construct those equations from a fresh situation, rather than always presenting the pair already formed. On another return, the equations may be supplied and the focus can be efficient execution. Both demands matter, but they are different.
Use contrast sparingly. A mixed set might place a proportional relationship beside a linear relationship with a fixed starting amount. The learner must decide which model fits. That prevents a strong technique from becoming an automatic response to every vaguely similar question.
A fresh result should also be completed properly. Units, solution sets and requested form remain part of mathematical performance. Maintaining a method while repeatedly neglecting its ending can preserve the wrong version of success.
When the student is consistently secure, reduce dedicated practice. A longer mixed paper may provide enough evidence for several maintenance decisions at once. Do not count a topic as practised simply because its name appears in the paper; inspect what the student actually did.
The goal is accessible, selectable mathematics. A learner should be able to find the method when the chapter label is absent, execute it accurately and recognise when another method is required.
14. Science maintenance should keep relationships alive
Science revision can drift towards lists of facts because they are easy to check quickly. Facts matter, but many examination tasks also require relationships: how a condition changes a process, how evidence supports a conclusion, or why a variable must be controlled.
Maintain a secure concept through a small changed-context task. If a student understands insulation, ask for a prediction in a different container setup. If they understand fair comparisons, ask which condition would make a new investigation difficult to interpret. Keep the task aligned with the learner’s curriculum rather than introducing irrelevant complexity.
One return can focus on recall of an essential definition. Another can ask for application. A later mixed section tests whether the student chooses the right concept. Rotating demands is more useful than repeating the same memorised sentence indefinitely.
When an answer weakens, locate the missing relationship. The student may still know the facts but omit the causal middle. That needs a different repair from forgetting the facts themselves. A brief oral explanation can help identify the distinction, but written examination performance still needs a written check.
Do not award maintenance success because the student agrees with the teacher’s explanation. The evidence is what they can produce before the explanation is supplied. Once the relationship is secure again, return it to maintenance instead of leaving it permanently in the intensive repair queue.
15. English and humanities maintenance should use new meaning
A learner may have a strong paragraph structure but lose flexibility when every practice response uses the same content. Maintenance should preserve the skill, not merely the familiar script.
For reading, use a short fresh passage and one carefully chosen question. Ask for evidence and a supported inference, or for the meaning of a phrase in context. A new passage creates a different access problem from rereading an answer already corrected in class.
For writing, a full essay is not always necessary. A new plan can test relevance and organisation. One paragraph can test the connection between claim, evidence and explanation. A short editing task can inspect a recurring language-control issue. Choose the sample according to the capability that was previously established.
Humanities knowledge can be maintained through a small argument task rather than a list of disconnected dates or facts. Ask which evidence would support a particular claim and why. Then occasionally change the command so that the learner must adapt selection rather than reproduce a memorised response.
These shorter checks do not replace whole-task practice. Students eventually need to sustain reading and writing across the full assessment. Maintenance uses small samples between those larger opportunities. The benefit is economy: it keeps important capabilities visible without demanding a complete essay every time the topic returns to the plan.
16. Avoid turning strong topics into a comfort refuge
Kai Kai likes maintenance because he likes being correct. Soon the short return becomes most of the evening. He chooses familiar questions, finishes quickly and reports a large amount of completed work. The weak topic remains untouched.
The solution is not to remove maintenance. Give it an endpoint. Select the task before starting, complete it independently, inspect the result and stop when the maintenance job is done. A secure result should release time for the harder work rather than justify another hour of comfortable repetition.
Watch for disguised repetition. Changing the numbers in an already familiar template can be useful for fluency, but after fluency is secure it may add little information. A brief fresh contrast may provide a better check than another long block of nearly identical work.
Parents should celebrate real success without letting success become the only acceptable feeling in study. A difficult repair block may produce fewer correct answers while doing valuable work. A maintenance block should usually be smaller because it starts from an established capability.
The balanced message is: “Keep what works, then spend the released time on what still needs help.” This avoids the false choice between only practising weaknesses and only practising strengths. Both extremes can keep a learner busy while limiting the improvement of the whole examination profile.
17. Avoid turning every return into a stressful test
Maintenance becomes unsustainable when each check feels like a verdict. Students may hide uncertainty, peek at answers or avoid the task because a mistake triggers adult disappointment. The evidence then becomes less trustworthy.
Keep ordinary maintenance low stakes. A wrong answer is information about what needs another look. It should not erase the student’s history of success or turn a temporary lapse into a permanent label.
Use a small sample, clear feedback and a specific next action. If the student cannot recall a formula, revisit its meaning, reconstruct it where appropriate and return later. If they know the formula but choose it in the wrong situation, inspect selection. Different failures should not all produce the same punishment of extra pages.
Do not require numerical scores for every return. “Independent and correct,” “needed a cue,” or “concept unclear” may be enough for the next decision. Formal scores become useful when the task and comparison conditions justify them.
The student should gradually participate in choosing the check and interpreting the result. That does not mean allowing only favourite questions. It means helping them understand why the sample exists and what it can show. Maintenance should build responsibility for keeping knowledge available, not dependence on a parent who constantly announces what might have been forgotten.
18. A changing deadline changes the maintenance decision
An examination next month and an examination tomorrow create different opportunities. With several weeks available, a student can investigate a weak maintenance result, teach the missing relationship and verify it after a delay. With one evening left, the decision must be more selective.
Near the examination, prioritise established skills that are important, vulnerable and realistic to refresh. Do not attempt to reactivate every chapter equally just because the entire syllabus feels urgent. A short retrieval check can identify what is already accessible and what needs a focused return.
Protect normal rest and the practical conditions needed for the paper. Adding maintenance should not simply extend an already overloaded schedule into the night. Remove low-value repetition before adding more time.
For back-to-back examinations, keep the current paper’s needs clear while preserving a small return route for later subjects. This is not the place to invent one universal subject order. The timetable, task complexity and learner’s current evidence should guide the sequence.
When the deadline has passed, reassess what deserves continued maintenance. Some knowledge supports the next course and should remain active. Some specialised examination details may no longer need the same attention. Maintenance is tied to future use, not a requirement to keep every school fact at peak accessibility forever.
19. Read the total and the topic profile together
A total score can rise while a former strength weakens, or remain flat while a difficult topic improves. Neither pattern should be interpreted from the total alone.
Open the paper and compare the relevant question families. Did the stronger topic actually appear at similar difficulty? Was the answer independent? Did timing change? Was the apparent decline a single arithmetic slip or a repeated loss of the underlying relationship?
The guide on passing totals that hide weak components is useful when strengths and weaknesses are compensating for each other. Its lesson is not to invent unofficial grades for every subsection, but to use available evidence to choose training.
For maintenance, ask whether the capabilities previously being protected remain usable. If they do, the maintenance dose may be adequate even when the overall score is noisy. If they do not, inspect whether the schedule failed to provide a return or whether the task has changed enough that a new level of learning is required.
Do not describe every score change as a causal effect of the latest timetable. School teaching, paper sampling and ordinary variation also matter. A cautious explanation can still be actionable: “This previously secure skill is less reliable on two fresh checks, so we will inspect and refresh it.” Certainty about the whole causal story is not required for a sensible small response.
20. Let the three learners choose different next actions
Alicia’s main problem was deletion. Once a topic was labelled strong, it disappeared from her plan. Her new approach preserves short independent returns and records a next-use date. She keeps most time on active repairs, but prior strengths no longer depend on being accidentally encountered again.
Tricia’s problem was overreaction. One imperfect maintenance response caused a full chapter restart. She now checks a fresh parallel item and the relevant explanation before deciding the scope of repair. Sometimes she needs a short correction. Sometimes a deeper lesson is justified. The size of the response follows the evidence.
Kai Kai’s problem was comfort. He used maintenance as permission to stay with easy work. His check now has a defined endpoint. A secure result sends him back to the difficult learning rather than into another set of familiar questions.
These are illustrative design changes, not reported outcomes from a trial. Their purpose is to show that “do more revision” is not a sufficient instruction. The same visible score problem can involve missing returns, excessive rebuilding or avoidance of difficult work.
A useful family discussion asks what the next small piece of evidence should be. It does not require a new app, a perfect spreadsheet or an elaborate colour scheme. A notebook entry and a well-chosen question can be enough to keep the learning plan connected to the student’s actual performance.
21. A maintenance record should help the student stop
The record needs only enough information to guide the next return: the capability, the last independent evidence, the current concern if any, and the next useful task. Avoid pages of historical scores that no longer affect a decision.
For example: “Percentage change: fresh mixed problem correct last Thursday; no current concern; include a reverse-percentage item next return.” Another entry might say: “Scientific explanation: evidence selected correctly but mechanism omitted; short repair before next maintenance check.” The two entries produce different actions.
Review the list occasionally. Retire redundant entries when ordinary schoolwork already supplies sufficient independent use. Combine overlapping checks when one task can legitimately sample several skills. Do not count the same answer as unlimited proof, but do avoid unnecessary duplication.
The record should also contain permission to stop. Once the planned maintenance evidence is obtained and secure, move to the next priority. Students should not feel obliged to keep checking until uncertainty is impossible. Educational evidence is always sampled; the goal is a reasonable decision, not omniscience.
As the learner becomes more independent, adults can reduce their involvement. The student learns to notice when a strong topic has not been used recently, choose an appropriate check and seek help only when the evidence calls for it. That is a more valuable outcome than perfect compliance with a timetable someone else maintains.
22. Three small checks, with three deliberately limited claims
For mathematics, ask the learner to solve 3(x − 2) = 18 and explain why the first transformation is valid. The solution is x = 8. A correct answer with a valid explanation supports current access to that elementary equation structure. It does not establish competence across all algebra. A later maintenance return might introduce a negative coefficient or a variable on both sides, provided those forms have already been taught.
For scientific reasoning, describe two otherwise comparable investigations that change different numbers of conditions. Ask which comparison makes it easier to interpret the effect of the intended variable and why. The important response identifies the competing changes rather than merely reciting “fair test.” Success supports that reasoning demand in the given scenario; it does not certify every practical-investigation skill.
For reading, provide this original sentence: “Although the room was ready, Sam checked the door twice before sitting down.” Ask the learner to distinguish what is stated from what might be inferred. Checking the door twice is stated. Uneasiness is a possible inference, not a directly stated fact. The learner should not invent a definite cause for that uneasiness without further evidence.
These examples illustrate task size, not a compulsory cross-subject quiz. Choose a check that matches the skill being maintained. Older or more advanced learners need appropriately demanding versions; younger learners need accessible wording and relevant prior teaching.
After each check, decide whether to continue maintenance, inspect an error or return to teaching. Do not turn one correct response into a broad mastery claim. The strength of maintenance lies in repeated, well-chosen opportunities over time, combined with ordinary coursework and larger performance checks. Small samples are useful because they inform small decisions.
23. Preserve progress while making progress
Weak topics deserve attention because they limit current performance. Strong topics deserve a smaller kind of attention because they are part of the capability the student is trying to carry into the examination.
The two jobs should not compete as moral choices. Working on a strength is not automatically avoidance. Working on a weakness is not automatically efficient. Ask what the task is doing and whether its size matches the evidence.
A balanced revision system has a way to build, consolidate, maintain and reopen. Skills move between those states as performance changes. No colour or label needs to become permanent.
The essential question is not “How much did I revise this week?” It is “What became more usable, and what remained usable while I was working on it?” That question includes both improvement and preservation.
Give the difficult work enough room to change. Give established capabilities enough return opportunities to stay available. Let independent performance determine when attention should increase or shrink. The result is not a timetable that treats all topics equally, but one that keeps the whole learner’s future performance in view.
Sources and further reading
Karpicke and Roediger, The Critical Importance of Retrieval for Learning examines continued retrieval after initial success in vocabulary learning. Cepeda and colleagues, Spacing Effects in Learning examines the interaction between review spacing and retention interval. The schedules and examples in this article are teaching proposals, not protocols tested by those studies.
Continue through the Complete Examination Craft Index or the Learning Runtime Hub.
