Maya has started well.

That matters.

She recognised the problem family without a topic label.

She selected a plausible method without the tutor naming it.

She made the first productive move.

She passed the Cold Start Test.

Five minutes later, however, the page is getting worse.

The algebra is longer.

The original structure is disappearing.

Each new line depends on another awkward fraction.

Maya has two possible responses.

The first is persistence.

“I started this way, so I must keep going.”

The second is panic-switching.

“This feels hard. I’ll abandon it and try something else.”

Neither is good enough.

The real skill is to ask:

Has this strategy stopped producing useful information—and if so, what evidence justifies changing it?

Maya looks back at the last line she still trusts.

She notices that expanding early destroyed a useful factorisation.

She rewinds.

Preserves the structure.

Chooses a different route.

The problem becomes shorter.

She has done something more advanced than “not giving up.”

She has replanned.

Independent learning requires not only starting a strategy, but monitoring whether it remains useful and changing course when the evidence says the current route is no longer worth pursuing.

That is the Replanning Trigger.

Quick Read: The Replanning Trigger in One Sentence

While using a strategy, monitor whether it is still reducing uncertainty, preserving important structure, producing checkable progress and matching the task; when several meaningful signals say it is not, return to the last trustworthy state, identify why the route failed, choose a better alternative and test the new plan.

This article owns strategy change.

The Cold Start Test asks whether the learner can begin.

The Replanning Trigger asks whether the learner can stop, inspect and choose again after beginning.

That is a higher-order form of independence.

1. Persistence Is Not a Strategy

“Never give up” sounds admirable.

In learning, it can be dangerously incomplete.

A learner can persist:

• with the wrong interpretation;
• with the wrong method;
• with an obsolete rule;
• with a representation that hides the structure;
• with a calculation that grows more fragile every line;
• with a writing plan that cannot support the claim;
• with a Science explanation missing its causal model.

Effort matters.

But effort should be governed by information.

Good persistence means continuing while the route is still learning-producing.

Good replanning means changing when the route has stopped earning its cost.

2. Switching Is Not Automatically Adaptive Either

The opposite learner abandons every method at the first friction.

One awkward line.

Switch.

One uncertain sentence.

Restart.

One moment of retrieval difficulty.

Open notes.

This can prevent productive struggle from doing its job.

The Replanning Trigger therefore sits between stubbornness and impulsive switching.

Do not change strategy because the work became difficult. Change strategy because the evidence says the current route is no longer a good use of effort.

3. Replanning Is a Metacognitive Control Function

Metacognition is often described through planning, monitoring and evaluation.

Replanning is what happens when monitoring changes the plan.

The learner notices:

this strategy is not working as expected.

Then evaluates why.

Then chooses a new route.

Current EEF guidance explicitly includes helping learners evaluate whether chosen strategies are working and adapt their approach when needed.

That small phrase—adapt their approach—is the centre of this article.

4. The Replanning Trigger Is Different From the Cold Start Test

Cold Start:

Can the learner choose a useful first route?

Replanning Trigger:

Can the learner recognise when that route has ceased to be useful and choose again?

A learner can be strong at one and weak at the other.

Ethan starts confidently but keeps pushing an inefficient method.

Hana starts slowly but changes course intelligently when evidence shifts.

Initiation and adaptation are different regulatory jobs.

5. The Replanning Trigger Is Different From the Intervention Threshold

The Intervention Threshold is tutor-side:

when should the tutor step in?

The Replanning Trigger is learner-side:

when should the learner change their own strategy before external help is needed?

A mature learner should increasingly perform the monitoring function that the tutor once performed from outside.

6. The Replanning Trigger Is Different From the Help Gradient

The Help Gradient governs external assistance.

The Replanning Trigger asks whether the learner can first generate a new internal plan.

The sequence can be:

current strategy fails → learner diagnoses why → learner tries alternative → only then, if needed, requests bounded help.

This protects independence without forcing endless unsupported struggle.

7. The Replanning Trigger Is Different From the Interference Test

The Interference Test asks whether two learned responses are competing.

The Replanning Trigger is broader.

A strategy can fail because:

• the task was misread;
• the representation is poor;
• the method is valid but inefficient;
• a prerequisite is missing;
• two rules interfere;
• an old heuristic is obsolete;
• execution drifted;
• the goal changed.

Interference is one possible diagnosis after the trigger fires.

8. The Replanning Trigger Is Different From the Supersession Test

The Supersession Test asks whether an old rule should remain, narrow, integrate or retire.

The Replanning Trigger asks what to do inside the current task when the current rule or route stops earning progress.

Sometimes replanning reveals supersession.

Sometimes it reveals only that a valid method was poorly chosen for this particular problem.

9. The Four Questions of Replanning

When a route feels wrong, ask four questions.

1. What evidence says the strategy is failing?

2. Is the strategy wrong, or is my execution wrong?

3. What is the last state I still trust?

4. What alternative has a reason to work better?

These questions prevent random restarting.

Replanning should preserve evidence from the failed route.

10. Signal One: The Strategy Stops Reducing Uncertainty

A productive strategy should teach you something as you use it.

You know more after line three than after line one.

You eliminate possibilities.

You expose structure.

You create a checkable intermediate result.

A dead route often does the opposite.

More work.

No new information.

This is the first major replanning signal.

Am I learning anything from the next step?

11. Signal Two: Complexity Grows Without Explanatory Gain

Some hard problems genuinely become complex.

Complexity alone is not failure.

But unnecessary complexity is a warning.

Maya’s factorised expression becomes twelve terms.

No new relation has appeared.

Nothing is easier to inspect.

The route has increased cognitive load without increasing control.

That is a valid reason to reconsider representation.

12. Signal Three: The Strategy Keeps Violating a Known Constraint

A method produces answers outside the permitted domain.

An English argument repeatedly drifts outside the question scope.

A Science explanation contradicts the observed result.

A geometry route depends on an unproven assumption.

When a route repeatedly collides with a task constraint, the learner should not merely push harder.

The route itself deserves review.

13. Signal Four: The Same Error Recurs After Local Repair

Ethan fixes one sign.

Two lines later, the same structural error appears again.

Local correction is not changing the system.

This suggests the problem may be upstream.

Representation.

Rule selection.

Hierarchy.

Replanning means moving upstream instead of repairing the same symptom repeatedly.

14. Signal Five: A Simpler Representation Exists but Is Being Ignored

The learner is doing heavy algebra on a relationship that a diagram would reveal immediately.

Or writing a long paragraph before deciding the claim.

Or memorising Science phrases when a particle model would organise the explanation.

A representation shift is often a replan, not a completely new solution.

Ask:

Would the problem become easier to reason about if I represented it differently?

15. Signal Six: The Strategy Cannot Explain Why It Should Work

“Why are you doing this?”

“Because that is what we did last time.”

That answer does not prove the method is wrong.

But it reveals weak strategic control.

A robust route usually has a task-based reason.

“I’m using elimination because the coefficients already align.”

“I’m comparing these paragraphs because the question asks about a shift.”

“I’m preserving factorised form because the roots matter.”

When the learner cannot state any reason, replanning may begin with reconstructing the goal.

16. Signal Seven: The Strategy No Longer Matches the Goal

Sometimes the route was sensible at the beginning.

Then new information changes the problem.

The learner discovers a hidden constraint.

An English counterargument changes the shape of the thesis.

A Science result contradicts the predicted mechanism.

Adaptive expertise includes responding to changing conditions rather than treating the original plan as sacred.

17. The Two-Signal Rule

Students often switch too early because one thing feels uncomfortable.

A useful practical safeguard is:

Do not abandon a plausible strategy on feeling alone. Look for at least two meaningful signals that the route is failing or one decisive contradiction.

Examples:

• complexity is growing and no new information appears;
• the same error recurs and local fixes do not change it;
• the method conflicts with a hard task constraint;
• the route requires an assumption known to be false.

This is an explanatory rule, not a validated threshold.

Its purpose is to slow impulsive switching while preserving adaptability.

18. Return to the Last Trustworthy State

When a route fails, do not erase everything.

Find the last point you still trust.

The correct interpretation.

The valid equation.

The evidence you know is relevant.

The confirmed observation.

Replan from there.

This preserves information earned by the failed route.

A failed strategy can still produce useful evidence.

19. Diagnose the Failure Before Selecting the Replacement

Do not switch from Method A to Method B merely because B is different.

Ask what made A poor.

• Did A hide structure?
• Did A create unstable arithmetic?
• Did A require knowledge I do not have?
• Did A conflict with a condition?
• Did A solve the wrong subproblem?
• Did A work but too slowly for the context?

The replacement should answer the diagnosed weakness.

Otherwise strategy switching becomes roulette.

20. Choose the Next Strategy by Contrast

Suppose substitution created fractions immediately.

A sensible alternative is elimination if it removes a variable cleanly.

Suppose chronological essay planning creates repetition.

A sensible alternative is claim-based organisation.

Suppose memorised Science phrasing cannot explain a novel situation.

A sensible alternative is rebuilding from the underlying model.

Good replanning chooses the next method because it corrects the previous route’s weakness.

21. Run a Cheap Test Before Fully Committing

Do not spend another ten minutes before learning whether the replacement is better.

Test it cheaply.

One line of algebra.

One diagram.

One paragraph outline.

One predicted consequence.

A good replan begins with a low-cost probe.

If the route starts producing better information, continue.

If not, inspect again.

22. The Replanning Ladder

A practical sequence is:

• 1 — Continue. The route is difficult but still informative.
• 2 — Inspect. Progress feels weaker; identify evidence.
• 3 — Local repair. Correct execution while preserving strategy.
• 4 — Reframe. Change representation while preserving the goal.
• 5 — Switch strategy. Choose a different method for the same goal.
• 6 — Return upstream. Revisit interpretation or assumptions.
• 7 — Escalate. Package the unresolved state and ask for bounded help.

This is an explanatory framework, not a validated scale.

Its job is to show that “keep going” and “give up” are not the only options.

23. Local Repair Before Global Switch

Maya’s method may be sound.

One sign is wrong.

Do not abandon elimination because of an arithmetic mistake.

Repair locally.

Similarly, Hana’s essay plan may be sound while one example is weak.

Replace the evidence, not the thesis.

Replanning should change the smallest level necessary.

This keeps good structure intact.

24. Representation Switch Before Method Switch

Sometimes the method is fine but the representation is poor.

A word problem is opaque until drawn.

A Science process is confusing until modelled as particles.

A writing argument is repetitive until reorganised as claims and evidence.

Try changing how the problem is seen before discarding the underlying method.

25. Goal Check Before Strategy Switch

A learner can optimise the wrong goal.

They are simplifying an expression beautifully.

The question asked for proof.

They are collecting quotations.

The question asked for an inference.

They are explaining temperature.

The question asked about heat transfer.

Before changing strategy, verify the target.

Sometimes the plan is failing because it is solving the wrong problem well.

26. Persistence Needs a Stop Rule

Students are often told to keep trying but not taught how to stop intelligently.

A useful stop rule might be:

If two genuinely different attempts produce no new information, return to the last trusted state and reframe the problem.

Or:

If the next likely step would merely repeat the same failure, do not repeat it.

These are operational rules, not universal laws.

Their value is making persistence inspectable.

27. Replanning Should Preserve the Failure Evidence

Do not erase the route that failed.

It may show:

• what assumption was made;
• what representation was chosen;
• where complexity exploded;
• what constraint was violated;
• which competitor was activated.

This becomes evidence for the Flight Recorder.

A failed route can become a future warning signal.

28. Replanning Should Produce a Better Question

The learner starts with:

“How do I solve this?”

After one failed route, the question can improve:

“Why does expansion make the structure harder to see?”

Or:

“Which quantity should remain invariant?”

Failure should refine the problem representation.

If the second attempt begins with exactly the same vague question, little has been learned from the first route.

29. Replanning and Adaptive Expertise

Adaptive expertise research distinguishes routine performance from the capacity to respond effectively when tasks or conditions change.

Routine expertise can be fast and accurate in familiar situations.

Adaptive expertise becomes visible when the familiar route is no longer sufficient.

The learner must reassess.

Reframe.

Experiment.

Choose differently.

The Replanning Trigger is one school-level mechanism that can help cultivate this flexibility.

30. Replanning and Expertise Reversal

Instruction that helps a novice can become unnecessary or even obstructive for a more knowledgeable learner.

A 2025 meta-analysis found a robust expertise-reversal pattern across many studies: lower-prior-knowledge learners benefited more from higher assistance, whereas higher-prior-knowledge learners tended to benefit more from lower assistance.

This matters for replanning.

Novices may need the tutor to model how strategy change works.

More advanced learners need space to monitor and change strategies themselves.

The same amount of tutor control is not optimal forever.

31. Primary 1: Replanning Can Be “Try Another Way” With Evidence

A young learner tries to read a word from the first letter only.

The guess does not fit the rest of the word.

Teach:

“Did your guess match all the letters? If not, try the sound groups.”

The child learns that evidence can stop a strategy.

Replanning begins very simply.

32. Primary 2: Replanning Means Returning to the Question

Ethan begins a word problem with addition.

The answer becomes larger than both quantities even though the story describes items being removed.

Instead of correcting him immediately, ask:

“Does the direction of your answer fit the story?”

The child learns to use reasonableness as a replanning signal.

33. Primary 3 and 4: Science Replanning Means Revising the Model

Jia Jun predicts one outcome.

The experiment gives another.

The wrong move is to rewrite the observation to fit the prediction.

The better move is:

“What part of my model must change to account for the result?”

Science is built around model revision.

Replanning is therefore deeply scientific.

34. Primary 5: Replanning Should Enter Homework Routines

A child gets stuck at home.

Before calling a parent:

• mark the last step that still makes sense;
• state what the current strategy was trying to achieve;
• try one different representation or simpler case;
• if still stuck, build an Escalation Packet.

This keeps home independence connected to intelligent help-seeking.

35. Primary 6: Replanning Is an Examination Time Skill

PSLE creates a cost to over-persistence.

A child can spend ten minutes forcing one dead route while easier marks remain elsewhere.

A useful examination replan can be:

• mark the question;
• write the last useful observation;
• move on;
• return later with a fresh representation.

Replanning includes time allocation, not only method choice.

36. Secondary 1: Replanning Protects Against “One Method for Everything”

Secondary 1 students often over-generalise newly learned methods.

Every ratio question gets the same template.

Every comprehension question gets a quotation.

Every Science answer gets memorised keywords.

Teach:

“What evidence would tell me this familiar method does not fit this problem?”

That question begins the move from routine to adaptive use.

37. Secondary 2: Replanning Should Compare Cost, Not Just Correctness

Two methods can both be valid.

One may be much more expensive.

Substitution creates fractions and long expressions.

Elimination removes a variable in one step.

Adaptive learners should notice this cost difference.

Replanning is not only about avoiding wrong methods.

It is also about replacing unnecessarily expensive ones.

38. Secondary 3: A-Math Replanning Begins With Structure Preservation

Maya expands too early.

The expression becomes harder to differentiate.

Her replanning trigger:

“Did my last move destroy a useful structure?”

If yes, return to the factorised state and choose again.

This one habit can protect differentiation, trigonometry, logarithms and algebraic manipulation.

39. Secondary 4: Replanning Must Become Fast

Near O-Level, the learner cannot run a ten-minute philosophical review every time a method feels awkward.

Compress the trigger.

Progress?

Constraint?

Structure?

Cost?

Alternative?

These five checks can run quickly.

If the route still has evidence behind it, persist.

If not, pivot.

40. JC: Replanning Becomes Model Selection

At JC, alternative strategies can reflect different models of the problem.

The learner may need to decide whether to:

• reason symbolically or graphically;
• use a limiting case;
• change variables;
• work from definition rather than formula;
• reframe an argument around a different principle.

Replanning at this level increasingly resembles expert problem solving:

not just changing procedures, but changing how the problem is represented.

41. English Comprehension: Replan When Evidence Does Not Support the Answer

Hana has an interpretation.

She cannot find textual evidence strong enough to support it.

That is a replanning signal.

Do not keep searching indefinitely for a quotation to rescue the preferred answer.

Ask:

“Is my interpretation wrong, or am I looking in the wrong part of the text?”

Then test both hypotheses.

42. English Writing: Replan the Argument Before Rewriting Every Sentence

An essay feels repetitive.

The student starts polishing wording.

Wrong level.

If the claims overlap, sentence editing will not repair the architecture.

Return upstream.

Replan claims.

Then rewrite.

Good replanning changes the level where the problem actually lives.

43. Mathematics: Replan When Algebra Stops Exposing Structure

A useful Mathematics route should usually reveal relationships.

If each line merely produces more symbols, ask:

“What am I trying to make visible?”

The answer may suggest:

• factorising;
• substituting;
• drawing;
• changing variables;
• using symmetry;
• testing a special case.

Replanning is often a search for a representation that carries more information per step.

44. Science: Replan When the Model Cannot Predict

A Science model should do more than reproduce a memorised phrase.

It should help predict what happens if conditions change.

If the student’s model cannot generate a prediction, or repeatedly predicts the wrong direction, that is evidence for revision.

Ask:

“What mechanism would have to be true for this result to occur?”

Then rebuild from mechanism.

45. Additional Mathematics: Replan at the Hierarchy Level

A-Math often punishes local repair when the hierarchy is wrong.

The student keeps correcting derivatives inside a product while the outer Product Rule structure is missing.

Stop repairing leaves.

Replan the tree.

Ask:

“What is the outermost operation governing everything below?”

This is a structural replan.

46. Replanning After AI Assistance

An AI tool suggests a route.

The route appears elegant but does not match the school method or task constraints.

The learner should not continue merely because the suggestion came from a powerful tool.

Apply the same replanning checks:

• what goal is this route serving?
• what assumptions does it make?
• does it preserve the required conditions?
• is it producing verifiable progress?
• can I explain why it should work?

Tool output should enter the learner’s monitoring system, not bypass it.

47. Replanning After Tutor Advice

A tutor recommends a method.

The student tries it.

It does not work as expected.

The learner should not assume:

“The tutor must be right, so I must be executing badly.”

Sometimes the tutor’s suggestion was provisional.

Good tuition permits evidence to update the plan.

That is intellectually healthier than authority replacing monitoring.

48. Replanning and the Escalation Packet

After one or two intelligent replans, the learner may still be stuck.

Now the Escalation Packet becomes richer.

Instead of:

“I don’t know.”

The learner can say:

“I tried substitution first. Fractions grew and gave no simpler relation. I returned to the original equations and tried elimination, but I still get inconsistent signs. I think the issue may be my setup rather than method choice.”

That is an expert-quality handoff.

49. Replanning and the Help Ledger

The Help Ledger can track:

strategy initiation — learner.

failure detection — tutor prompted.

alternative generation — learner.

This tells us exactly what still needs to migrate.

The learner may already know alternative strategies but still depend on somebody else to say:

“This is not working.”

That monitoring function becomes the next target.

50. Replanning and Prompt Migration

Useful replanning prompts include:

• Am I learning anything from the next step?
• What is the last line I trust?
• Did my representation make the structure clearer or harder?
• Am I repairing the same error repeatedly?
• What constraint is my plan violating?
• What alternative directly fixes that weakness?

At first the tutor asks.

Later the learner asks.

This is Prompt Migration applied to strategy control.

51. Replanning and the Diagnostic Probe

Two explanations remain.

Bad strategy?

Bad execution?

Run a Diagnostic Probe.

Give a simplified version where execution demand is low.

If the learner still chooses poorly, strategy selection is implicated.

If selection is good but errors remain, execution may be the owner.

Replanning becomes smarter when causes are discriminated rather than guessed.

52. Replanning and the Evidence Threshold

One failed strategy does not justify rewriting the whole programme.

Repeated failure to detect dead routes across subjects may.

The Evidence Threshold asks when there is enough evidence to change the teaching plan.

A pattern of tutor-triggered replanning can justify explicit monitoring training.

53. Replanning and the Transfer Gate

The learner replans successfully in tuition.

Good.

Now test:

• different wording;
• mixed methods;
• delay;
• time pressure;
• school paper.

The Transfer Gate asks whether strategic flexibility travels.

Adaptive expertise is not one successful pivot under coaching.

It is a capability that survives new conditions.

54. Replanning and the Maintenance Window

A skill returns after a long gap.

The learner begins with an old inefficient route.

Can they detect and change it independently?

A good Maintenance Window should sometimes test adaptive control, not only recall of the final procedure.

Stable expertise includes knowing what to do when the first route is no longer ideal.

55. Replanning and Family Life

At home, parents often enter when a child’s first strategy fails.

“Do it this way.”

That may solve the homework.

It may also remove the chance to practise replanning.

A better parent response can be:

“Show me where your current way stopped helping. What could you change without me giving you the method?”

If the child cannot generate an alternative, package the problem for tuition.

The parent preserves the learning job without becoming the replacement strategist.

56. The Parent’s Replanning Test

• Does my child keep repeating the same failed move?
• Can they identify what the move was trying to achieve?
• Can they name evidence that it is not working?
• Can they return to the last trustworthy state?
• Can they suggest one alternative?
• Do I routinely supply the alternative before they try?
• Can I ask a monitoring question instead of giving a method?
• When should this move to the tutor?

The goal is not parental silence.

It is preserving the learner’s strategy-control job where possible.

57. The Tutor’s Replanning Test

• Did the learner choose the first strategy independently?
• What evidence shows the route is weakening?
• Has the learner noticed that evidence?
• Can they distinguish strategy failure from execution failure?
• Can they identify the last trustworthy state?
• Can they generate at least one justified alternative?
• Can they cheaply test the alternative?
• Do they switch too early?
• Do they persist too long?
• Which monitoring prompt should migrate inward?
• Does replanning survive mixed and delayed tasks?

This makes “flexibility” observable.

58. The Student’s Replanning Test

An older learner can ask:

• What is my current strategy trying to accomplish?
• What evidence says it is working?
• What evidence says it is not?
• Am I learning anything from the next step?
• Am I fixing the same symptom repeatedly?
• What is the last state I trust?
• What alternative directly addresses the weakness?
• What is the cheapest test of that alternative?
• If both routes fail, what specific help do I need?

These are questions of strategic control.

59. The Three-State Replanning Record

A lightweight record can use three states.

Start state:
What strategy was chosen and why?

Trigger state:
What evidence caused reconsideration?

Replan state:
What changed and why should the new route be better?

Example:

Start — substitution because one variable looked isolated.

Trigger — fractions grew, no variable was simplified, arithmetic risk increased.

Replan — elimination because coefficients align after one multiplication.

This makes strategic learning visible.

60. The “Why I Switched” Rule

After a learner changes strategy, ask for one sentence:

“I switched because…”

Good answers:

“The first method was valid but created fractions that made error more likely.”

“My interpretation had no evidence in the text.”

“The model could not explain the observed result.”

Poor answer:

“I felt like trying something else.”

The sentence forces the switch to have an evidence basis.

61. Replanning Can Be Too Frequent

Some learners become strategic tourists.

They visit every method.

Commit to none.

This creates shallow exploration.

The tutor should teach commitment windows:

“Give the plausible route enough steps to generate evidence before judging it.”

Adaptive expertise is not constant change.

It is justified change.

62. Replanning Can Be Too Late

Other learners attach identity to the original method.

“I always do it this way.”

They continue because abandoning the method feels like admitting failure.

Teach:

Changing strategy because of evidence is not failure. It is higher-quality control.

The goal is not loyalty to the first plan.

The goal is solving and learning well.

63. Replanning Protects Against “More Practice” as the Default Repair

A student repeatedly chooses the wrong strategy.

The response is fifty more questions.

That may simply automate the wrong initiation and weak monitoring.

Sometimes the learner does not need more repetitions.

They need better decision rules and replanning triggers.

Volume cannot repair every control problem.

64. Replanning Protects Against Tutor Takeover

The tutor sees the route failing before the student does.

Temptation:

“Stop. Do it this way.”

Sometimes necessary.

But if used every time, the student never practises failure detection.

When safe, ask:

“What is your current method giving you that you did not already know?”

Let the learner discover the need to replan.

65. Replanning Protects Against Random Experimentation

Trial-and-error can be productive.

Random trial-and-error wastes evidence.

Each attempt should answer something.

“If I preserve factorisation, does the derivative structure become clearer?”

“If I reorganise by claim rather than chronology, does repetition reduce?”

“If I use the particle model, can I predict the changed condition?”

Replanning should turn attempts into experiments.

66. Replanning Protects Against False Grit

Grit without monitoring can become expensive stubbornness.

True perseverance is not:

“I will use this method forever.”

It is:

“I will stay committed to the learning goal while changing methods when evidence justifies it.”

That is a more mature form of persistence.

67. Replanning Protects Against False Flexibility

Constant switching can look creative.

Sometimes it is avoidance of depth.

A learner who changes methods every two lines may never learn what any method actually reveals.

Adaptive flexibility requires a stable enough attempt to generate evidence.

Then change.

Not before.

68. The Replanning Trigger in a Three-Student Class

Maya expands and later replans.

Hana preserves the first route because it is still working.

Jia Jun switches too early.

The tutor can compare these behaviours without giving all three the same command.

This makes strategic control visible.

Small-group tuition can therefore teach not only methods, but method governance.

69. Back to Maya

Maya looks at the page again.

Her first route was not foolish.

It was plausible.

That matters.

The goal is not to teach students to predict the perfect method instantly.

Real problems do not always allow that.

The deeper capability is:

choose intelligently.

observe honestly.

change deliberately.

She returns to the factorised expression.

Tests the alternative in one line.

Sees that the structure is now clearer.

Continues.

The tutor does not rescue her.

The tutor watches a more advanced thing happen.

Maya is governing her own strategy.

---

The Replanning Trigger in One Page

1. Persistence is not enough.
Continue while the route is useful, not merely because it was chosen first.

2. Switching is not automatically adaptive.
Do not abandon a plausible strategy at the first discomfort.

3. Monitor progress.
Ask whether the strategy is reducing uncertainty and exposing structure.

4. Watch for meaningful signals.
Growing useless complexity, repeated errors, violated constraints and absent progress all matter.

5. Distinguish strategy failure from execution failure.
Repair locally before replacing globally.

6. Return to the last trustworthy state.
Do not throw away information earned by the first attempt.

7. Diagnose why the first route failed.
The alternative should correct that weakness.

8. Test the new strategy cheaply.
Use one line, one example or one representation before fully committing.

9. Use stop rules.
Do not repeat the same failed move when the next step is unlikely to add information.

10. Replan at the right level.
Sometimes change execution, sometimes representation, sometimes strategy, sometimes interpretation.

11. Preserve the failed route.
It contains diagnostic evidence.

12. Explain why you switched.
Evidence-based replanning is different from random experimentation.

13. Let monitoring prompts migrate inward.
The learner should increasingly detect dead routes before the tutor does.

14. Keep the goal stable while methods change.
Adaptive persistence means loyalty to the learning goal, not loyalty to the first plan.

15. Remember the destination.
Independent learners should be able to start, monitor, replan, recover and escalate without waiting for another person to govern every strategic decision.

Where This Fits in eduKatePunggol

• How Tuition Works at eduKatePunggol — the broad tuition relationship.
• The Weak Link — identify the first limiting mechanism.
• The Learning Simulator — safe conditions for testing alternative strategies.
• The Sparring Partner — intelligent resistance that forces strategy evaluation.
• The Flight Recorder — preserve failed routes and strategy-switch histories.
• The Learning Dispatcher — route problems that remain unresolved after internal replanning.
• The Exit Ramp — transfer strategy governance toward the learner.
• The Priority Queue — prioritise repeated replanning failures with high transfer value.
• The Intervention Threshold — decide when tutor intervention should follow failed learner replanning.
• The Translation Layer — identify whether the route failed because the task itself was misread.
• The Transfer Gate — test whether adaptive strategy control survives new contexts.
• The Diagnostic Probe — separate strategy failure from execution failure.
• The Evidence Threshold — decide when repeated poor strategy control warrants a plan change.
• The Counterfactual Check — interpret improved adaptive control without overclaiming causality.
• The Maintenance Window — verify that strategy monitoring remains available after delay.
• The Reactivation Cost — distinguish a cooled strategy from a genuinely poor strategy choice.
• The Interference Test — identify when a competing learned response drives the failing route.
• The Supersession Test — decide whether the old strategy should remain or retire.
• The Escalation Packet — hand over a well-diagnosed problem after internal replanning is exhausted.
• The Help Gradient — choose the smallest useful external assistance after replanning fails.
• The Help Ledger — identify whether the tutor still supplies failure detection or alternative generation.
• Prompt Migration — move strategy-monitoring questions from tutor to learner.
• The Cold Start Test — test whether the learner can choose and begin a first strategy independently.

Batch 06 now has two advanced independence mechanisms:

• The Cold Start Test — can the learner begin without external ignition?
• The Replanning Trigger — can the learner detect when the chosen route has stopped earning progress and deliberately choose another?

Together:

Recognise → start → monitor → detect mismatch → return to trusted state → replan → test alternative → continue or escalate.

---

Research Foundations and Evidence Boundaries

The Replanning Trigger is an explanatory tuition framework, not a validated diagnostic instrument. Its signals, two-signal rule and ladder are practical teaching models rather than formal psychometric scales. The evidence is stronger for the underlying principles: metacognitive monitoring, evaluating whether strategies are working, adapting strategy when conditions change, adaptive expertise and matching instructional assistance to learner knowledge.

• Education Endowment Foundation — Metacognition and Self-Regulation: Developing Independent Learners (24 August 2026). EEF highlights planning, monitoring and evaluation and explicitly recommends helping learners judge whether chosen strategies are working and adapt their approach when needed.
• Bohle Carbonell et al. — How Experts Deal With Novel Situations: A Review of Adaptive Expertise, Educational Research Review (2014). The review links adaptive expertise to flexible problem solving, knowledge representation and learning environments that allow learners responsibility for developing solution strategies.
• Pelgrim et al. — Professionals’ Adaptive Expertise and Adaptive Performance in Educational and Workplace Settings: An Overview of Reviews (2022). The overview describes adaptive performance as visible response to changed or novel task conditions and emphasises the need to find new solutions when routine responses are insufficient.
• Adaptive Expertise Development During Work-Based Learning in Higher Education: A Realist Review (2025). The review identifies shifts in thinking and working in response to challenging situations as a recurring mechanism associated with adaptive expertise development.
• A Cornerstone of Adaptivity — A Meta-Analysis of the Expertise Reversal Effect, Learning and Instruction (2025). Across 60 experimental studies, lower-prior-knowledge learners benefited more from higher assistance while higher-prior-knowledge learners benefited more from lower assistance, reinforcing the need to adjust instructional support as expertise changes.

These sources do not imply that students should constantly switch methods or that every difficult strategy is wrong.

A more defensible conclusion is:

Independent learners benefit from being taught not only how to select strategies, but how to monitor whether those strategies remain effective, interpret evidence of mismatch, and adapt their approach deliberately when conditions warrant change.

That is the Replanning Trigger.

Not stubbornness.

Not random switching.

Evidence-guided strategic change.