The Tutor Handbook Vol No.0012 | The First Review — How a Tutor Decides Whether to Continue, Advance, Reclassify or Stop a Learning Route

The Tutor Handbook · Volume 0012 · Series ID THB-0012

A learning route should not continue merely because it has begun.

That sounds obvious.

In practice, it is one of the hardest disciplines in tutoring.

A tutor finds a weakness. A plan is built. Worksheets are selected. The parent settles into a routine. The student becomes familiar with the lesson structure. Weeks pass.

Then something important happens.

The evidence changes.

The learner may be ready for something harder. The original weak link may have been repaired. A different bottleneck may have emerged. The route may be producing little change. The tutor may have classified the problem incorrectly. The current class size may no longer fit. The student may need less support. Or the problem may sit outside ordinary tuition.

At that point, a tutor has four serious choices:

Continue. Advance. Reclassify. Stop.

The first review is where tutoring proves whether it can read its own evidence.

This is Volume 0012 of The Tutor Handbook, eduKate Sengkang’s long-form practical series on the real jobs inside tutoring. Volume 0009, The First Lesson, established a first working map. Volume 0010, The Second Lesson, tested the first hypothesis. Volume 0011, The First Month, used repeated lessons to build a stable teaching route.

This volume owns the next decision:

Once a route has enough evidence behind it, how does the tutor decide whether it deserves to continue?

The wider tutor ladder remains owned by the Tutor Classification Model by eduKateSG and the Tutorial Class Control Tower. General route sequencing remains owned by Tutor Handbook Vol No.0005 | Route Designer. Tutor–student fit remains owned by Volume 0008. This page is narrower: it is the first formal decision gate after enough repeated evidence has accumulated to make continuation itself a hypothesis worth testing.

Quick Read

• The first review asks whether the current learning route still deserves to continue.
• Do not review attendance alone. Review learner change under declared support conditions.
• Continue when the route still targets the dominant bottleneck and evidence shows meaningful movement.
• Advance when the current finish condition has been met and the learner is ready for harder, more mixed, more independent or more performance-like conditions.
• Reclassify when the original explanation no longer fits the evidence, or when the dominant tutoring job has changed.
• Stop a route when it has completed its job, produces little useful change, creates increasing dependence, is structurally mismatched, or the problem belongs outside the present tutoring arrangement.
• One unusually good or bad lesson should rarely decide the review by itself. Look for repeated patterns and changed-condition evidence.
• The tutor should distinguish the learner’s improvement from improvement in the work product created with tutor help.
• A review should end with a decision, the evidence supporting it, the next condition to test and what remains uncertain.
• The best review can conclude that the learner needs less tuition.

1. A Route Is a Working Theory, Not a Promise

When a tutor chooses a route, they are making a claim.

The claim is roughly:

If we work on this bottleneck, using this support, in this sequence, the learner should become able to do something important more independently.

That is a hypothesis.

The review asks whether the evidence still supports it.

A weak tutoring system treats the plan as an instruction to follow.

A stronger tutoring system treats the plan as something reality is allowed to correct.

2. The First Review Is Not a Parent Update With a New Name

Parents should receive updates.

The first review is more demanding.

An update says what happened.

A review decides what should happen next.

Update	Review
We covered fractions and ratio.	Ratio representation is now stable enough to move from blocked to mixed practice.
She is improving.	Prompt dependence has reduced from full modelling to one relationship cue across three changed tasks.
He still makes careless errors.	Most remaining errors occur under time pressure after long algebraic chains; untimed accuracy is now stable.
We will keep practising.	The present route will continue for two more cycles because the finish condition has not yet been met.

The review is a decision instrument.

3. The First Review Needs a Reference Point

Improvement has meaning only relative to something.

That something can be:

• the first-lesson baseline;
• the learner’s original failure signature;
• the support level initially required;
• a school performance before the intervention;
• a delayed return from earlier work;
• the route’s explicit finish condition.

If the tutor cannot say what the present learner is being compared with, “progress” becomes impressionistic.

The first review should therefore begin by reconstructing the original route.

4. Reconstruct the Original Claim Before Judging It

Ask four questions.

1. What did we think the main problem was?
2. What change did we expect if we were right?
3. What support did we plan to use?
4. What evidence did we say would count as readiness to move on?

If those questions were never answered, the review has discovered an earlier weakness in the tutoring design.

You cannot fairly review a route whose intended job was never defined.

5. The Review Should Ask What Changed in the Learner

Not what changed in the worksheet.

Not what changed in the tutor’s explanation.

Not how many chapters were covered.

What can the learner now do?

• Start more independently?
• Choose the method?
• Retrieve after a delay?
• Explain the mechanism?
• Use feedback?
• Recognise the same structure when the surface changes?
• Check a recurring error?
• Recover from difficulty?
• Perform under time?
• Decide what to practise next?

The route exists to change learner capability.

The review should remain anchored there.

6. Improvement Under Full Tutor Support Is Not the Same as Independent Improvement

A student may look dramatically stronger after a month because the tutor has become better at helping them.

That is useful.

It is not yet the whole learner receipt.

The review should separate at least three states:

State	Question
Supported performance	What can the learner do with the present tutor help?
Reduced-support performance	What survives when prompts are intentionally reduced?
Independent return	What survives later without the tutor beside the learner?

A route is strongest when all three are improving in the intended direction.

7. The Review Should Use More Than One Kind of Receipt

A score is one receipt.

It should not be the only one.

• accuracy;
• method selection;
• time to start;
• prompt count;
• error recurrence;
• delayed retrieval;
• transfer;
• completion under realistic conditions;
• quality of self-correction;
• quality of questions asked;
• ability to explain the next learning priority.

Different routes need different receipts.

A Drill Builder route should not be judged primarily by student enjoyment.

A Performance Coach route should not be judged primarily by untimed worksheet accuracy.

A Route Designer route should not be judged by the number of pages completed.

The receipt should fit the job.

8. The Review Should Compare Conditions

Performance changes when conditions change.

The review should know which conditions matter.

• timed versus untimed;
• single-topic versus mixed;
• familiar versus changed surface;
• with notes versus closed book;
• with tutor prompts versus without;
• immediate versus delayed;
• tuition-generated versus school-generated task.

If a learner succeeds only under one narrow condition, the route may not yet be ready to advance.

If the learner succeeds across several relevant conditions, the current route may have done enough.

9. One Good Week Should Not Erase Three Weak Weeks

Humans overweight recent evidence.

The latest lesson feels vivid.

A student performs brilliantly today and the tutor concludes the problem is solved.

Or the student has one difficult session and the tutor throws away a route that had been working.

The first review should ask:

Is this a new pattern, or one local performance?

Repeated evidence deserves more weight than one dramatic moment unless that moment reveals a genuinely new condition or safety concern.

10. Continue: When the Route Still Deserves More Time

Continue is not the passive choice.

It should mean:

The present route still targets the dominant bottleneck, meaningful progress is visible, the finish condition has not yet been met, and another cycle is likely to produce useful information or capability.

Continue when:

• the same failure signature is reducing but not yet stable;
• prompt dependence is falling;
• delayed return is improving;
• transfer has begun but remains inconsistent;
• current school demands still make the route relevant;
• the learner can explain the purpose of the current work;
• the route remains sustainable.

Continuation should still have a next review condition.

11. Continue Does Not Mean Repeat the Same Thing Unchanged

A route can continue while the task changes.

For example:

same learner job → less support → greater variety → longer delay → more independent condition.

The route is still working on the same bottleneck.

The training environment evolves because the learner is changing.

12. Advance: When the Present Route Has Done Its Job

Advance means the current finish condition has been met strongly enough that staying at the same level creates diminishing value.

Advancement can mean:

• less prompting;
• more difficult questions;
• mixed topics;
• changed representations;
• longer delay;
• greater speed demand;
• more authentic examination conditions;
• more responsibility for planning and checking;
• movement from Repair to Alignment or Frontier.

Advance does not necessarily mean a harder chapter.

Sometimes the next difficulty is independence.

13. The Independence Advance Is Often More Important Than the Content Advance

A student has learned the method.

The easiest next move is to teach a harder method.

But the better next move may be to remove the tutor’s prompt.

Can the learner:

• recognise when to use the method?
• start without a cue?
• check without being reminded?
• decide when they are finished?
• return to the work later?

The Education Endowment Foundation’s August 2026 material on metacognition and self-regulation emphasises planning, monitoring and evaluating learning as processes that can become increasingly internal over time.

The first review should therefore ask whether the learner is ready for a transfer of control, not only a transfer to harder content.

14. Reclassify: When the Original Tutor Job Has Changed

Reclassification is one of the most important outcomes because tutoring problems change as they are repaired.

A learner may begin with Class 3 Diagnostic Tutor work.

Once the weak link is clear, the dominant job may become Class 1 explanation or Class 2 fluency.

Later the learner may need Class 4 sequencing.

Closer to examinations, Class 5 performance may dominate.

The tutor does not need to change person every time the class changes.

The tutor does need to change function.

15. Reclassify When the Evidence Contradicts the Original Story

Suppose a learner was classified as weak in content.

Over several weeks, untimed independent reattempts are strong.

The failure appears almost exclusively under mixed timed papers.

The route should shift toward performance conversion.

Or suppose the learner was thought to have a performance problem, but untimed reattempts remain weak and explanations reveal conceptual confusion.

Return upstream.

Reclassification is what happens when evidence becomes more important than the tutor’s first impression.

16. Reclassify the Tuition Mode When the Learner Moves

The tutor should also review the mode described in eduKateSG’s Three Modes of Tuition.

Mode	Review question
Repair	Is the weak foundation still the dominant bottleneck?
Alignment	Is the learner now stable enough that staying connected to school should dominate?
Frontier	Has the learner earned enough stability for deeper transfer, justification and stretch?

A learner should not remain permanently in Repair simply because tuition began there.

17. Stop: The Most Misunderstood Review Outcome

Stopping is often treated as failure.

Sometimes it is the clearest sign of success.

Stop the current route when:

• the finish condition has been met;
• the route no longer targets the dominant problem;
• repeated evidence shows little useful change;
• the route is creating increasing support dependence;
• the tutorial structure is a poor fit;
• the learner has moved beyond the tutor’s expertise;
• the problem belongs outside ordinary tuition;
• the learner no longer needs this level of tutoring.

Stopping a route is not the same as abandoning the learner.

It is refusing to continue an intervention after its justification has expired.

18. Stop Because the Route Worked

This outcome deserves more attention.

Alicia no longer needs the relationship cue.

She represents familiar and unfamiliar word problems independently.

She retrieves the routine after a delay.

The repair route is finished.

Do not keep “repairing word problems” because that is what tuition has always done.

Either advance to another job or reduce tuition.

19. Stop Because the Route Is Not Working

A route has had enough time, practice and appropriate support.

The same error remains unchanged.

The same prompt remains necessary.

Transfer does not improve.

Another month of identical intervention is not persistence.

It is failure to update.

Stop the route and return to diagnosis.

20. Stop Because the Tutor Is Becoming the Missing Function

The learner appears stronger.

But every improvement requires more tutor regulation.

The tutor now:

• chooses every task;
• initiates every session;
• identifies every error;
• decides every return;
• provides every performance reminder;
• calms every uncertainty.

The tutoring system may be getting better while the learner’s independent system is not.

The review should reduce support, redesign the route or stop the present arrangement if dependence keeps growing.

21. Stop Because the Problem Belongs Somewhere Else

A tutor may observe persistent distress, health concerns, family crises, welfare issues or specialist learning needs that ordinary subject tuition cannot resolve.

The tutor can record educational observations.

The tutor should not convert every concern into a tuition diagnosis.

Stopping or narrowing tuition can be responsible when the correct next owner is school support, a health professional, a specialist service or simply a more sustainable family schedule.

22. The Review Should Protect Against Sunk-Cost Thinking

“We have already spent three months on this.”

That statement explains the past.

It does not justify the next month.

The correct review question is:

Given what we now know, is another cycle of this route the best next use of the learner’s time?

Past effort is not permission for future repetition.

23. The Review Should Protect Against Novelty Bias Too

Changing the plan feels active.

That can make reclassification attractive even when the present route is working.

A learner who is steadily reducing prompt dependence may not need a new system just because progress is no longer dramatic.

The review should change only what evidence justifies.

24. The Green–Amber–Red Review

A simple review can use three colours.

Signal	Meaning	Likely decision
Green	The intended capability is improving and the route still fits	Continue or advance
Amber	Some movement exists but the mechanism, fit or transfer remains uncertain	Continue briefly with a discriminating test or reclassify
Red	Little useful change, increasing dependence, clear mismatch or inappropriate scope	Reclassify or stop

The colours describe the tutoring arrangement.

They are not labels for the learner.

25. Green Is Not “Everything Is Perfect”

A green route can still contain mistakes.

It means the direction is defensible.

The learner may still be weak.

But the weakness is becoming smaller, more independent or more transferable in the expected way.

Continue because the mechanism is moving.

26. Amber Is Where the Best Tutors Become More Curious

Amber means the route is not clearly wrong.

It is not yet clearly strong either.

Examples:

• accuracy improves but prompt dependence does not;
• tuition worksheets improve but school transfer does not;
• the learner is faster but makes more errors;
• one topic improves while a shared prerequisite remains unstable;
• marks rise but the assessment was much more familiar.

Amber should produce a discriminating test.

Not vague waiting.

27. Red Does Not Mean the Learner Failed

Red means the current arrangement is no longer justified strongly enough.

The route may be wrong.

The tutor function may be wrong.

The class structure may be wrong.

The target may be unrealistic for the present time horizon.

The learner may need a different kind of support.

The student is not the colour.

28. The Review Should Include the Student

Do not review the learner entirely above the learner’s head.

• What feels easier now?
• What still feels uncertain?
• What do you do differently when stuck?
• Which tutor prompt do you need less often?
• What do you think we should continue?
• What feels repetitive without being useful?
• What do you want to try next?

The learner’s judgement may be incomplete.

It is still part of the evidence.

The EEF’s August 2026 guidance on developing independent learners describes feedback as more useful when it helps learners understand what success looks like and what specific next steps they can take, supporting greater ownership over time.

29. The Review Should Include the Parent—but Keep the Language Operational

A parent review should not become a catalogue of personal traits.

Prefer:

She now starts most ratio questions independently but still needs one cue when the reference quantity is hidden in comparative language.

over:

She is still weak and lacks confidence.

The first statement can guide the next route.

The second risks becoming a story about the child.

30. The Review Should Compare Tutor Prediction With Reality

At the beginning of the route, the tutor should have predicted something.

Perhaps:

If we repair sign handling, errors across functions and trigonometry should fall.

Now compare.

Did the predicted downstream change occur?

If yes, confidence in the model rises.

If no, the review should ask why.

This is one way tutors become better calibrated over time rather than merely more experienced.

31. Tutor Experience Should Improve Prediction, Not Harden Assumptions

Experienced tutors recognise patterns quickly.

That is an advantage.

It can also create premature certainty.

“I have seen this a hundred times.”

Perhaps.

The hundred-and-first learner still deserves direct evidence.

The first review should make experience accountable to what actually happened.

32. Review the Smallest Useful Prompt

Prompt reduction is one of the cleanest indicators of growing independence.

At the start:

The tutor models the entire first example.

Later:

The learner proceeds after one cue.

Later still:

The learner catches the difficulty independently.

That sequence can justify advancement even before the headline school mark changes dramatically.

33. Review the Return Path

What came back after a week?

What disappeared?

A route that works only inside the teaching session is fragile.

Return evidence helps distinguish immediate familiarity from durable availability.

eduKate Sengkang’s How We Know Learning Has Really Held develops this larger idea.

34. Review Transfer Before Declaring a Route Complete

The learner solves the exact kind of question practised.

Now change the surface.

Change the numbers.

Change the representation.

Remove the topic label.

Place the method beside a plausible neighbour.

If the capability survives, advancement becomes more defensible.

If it collapses, continue or reclassify.

35. Review the School Return

Tuition can create a local success that never returns to school.

The first review should inspect:

• new schoolwork;
• teacher comments;
• recent quizzes;
• homework completed without tutor presence;
• how the learner handles the school’s notation and question style.

The EEF’s evidence on one-to-one and small-group tuition continues to emphasise the importance of targeting learner needs and maintaining links with ordinary classroom teaching.

A tutoring route should eventually prove itself outside the tutor’s own materials.

36. Review the Workload

A route can be academically sound and practically unsustainable.

The learner has school, other subjects, CCAs, sleep, family life and ordinary human limits.

Ask:

• Is the learner completing the route without chronic late-night work?
• Is homework volume crowding out recovery?
• Are too many weaknesses active simultaneously?
• Is the route competing destructively with school?

An unsustainable route should be narrowed even if every individual task is educationally defensible.

37. Review the Tutorial Size

Tutor–student fit can change without changing the tutor.

A learner who began in Micro tuition because high visibility was necessary may now be ready for Meso conditions that require more self-direction.

A learner who is disappearing inside a group may need more visibility temporarily.

The review should ask:

Does the current tutorial size reveal and strengthen the behaviour we now need?

38. Small Groups Create a Useful Review Signal: What Happens When Tutor Attention Moves?

In a three-pax class, the tutor cannot regulate every learner continuously.

That creates an independence test.

When the tutor turns to another student:

• does Alicia continue?
• does Beatrice know what to do next?
• does Ciara wait?
• does Denise check?
• does Emily extend?
• does Faith abandon the task?

The learner’s behaviour during tutor absence can be more informative than behaviour during direct tutor attention.

39. One-to-One Reviews Must Deliberately Create Tutor Absence

One-to-one tutoring has extraordinary visibility.

It can therefore hide dependence exceptionally well.

During a review cycle, the tutor should deliberately remain silent for bounded independent tasks.

Otherwise the review measures the pair more than the learner.

40. Alicia: Continue

Alicia began with weak representation in Mathematics word problems.

At review:

• arithmetic remains strong;
• familiar relationship structures are now represented independently;
• one short cue still restores progress on dense comparison language;
• the same repair survives after a week;
• unfamiliar mixed problems remain inconsistent.

Decision: Continue.

Not because Alicia is still “weak at word problems”.

Because the route is producing the predicted change and has not yet met the transfer finish condition.

41. Beatrice: Advance

Beatrice’s comprehension route targeted question-demand paraphrasing and evidence selection.

At review:

• she identifies causal, attitude and purpose demands accurately;
• she selects evidence independently;
• the skill survives new passages;
• she no longer needs the tutor to name the relationship.

Decision: Advance.

The next challenge is answer compression and performance timing, not more blocked question-type practice.

42. Ciara: Reclassify

Ciara’s first route assumed conceptual weakness.

Repeated sessions show something different.

• untimed understanding is strong;
• she reattempts missed questions successfully later;
• performance collapses mainly in mixed timed sets;
• she spends too long on early difficult questions.

Decision: Reclassify.

The dominant job shifts from explanation toward Class 5 Performance Coaching.

43. Denise: Stop the Present Route

Denise has completed a targeted algebraic sign-repair route.

• sign-loss errors have fallen sharply;
• the change appears across equations, functions and trigonometry;
• she catches most remaining slips herself;
• the repair survives mixed work.

Decision: Stop the route because it worked.

Do not keep giving sign drills out of habit.

Move forward.

44. Emily: Stop Some Tuition

Emily is now strong, organised and independently stable.

The tutor can still stretch her.

But the review asks whether weekly frequency still pays enough educational rent.

Decision: reduce frequency, move to periodic frontier or performance review, or pause.

Successful tutoring should contain a route toward needing less of it.

45. Faith: Amber—Run One Discriminating Test

Faith’s Science explanations have improved strongly on tutor-created questions.

School-transfer evidence is still weak.

Decision: Amber.

Before continuing another month unchanged, use a school-generated or unfamiliar task under reduced support.

The next test should decide whether to continue or advance.

46. The Review Should Ask What the Tutor Is Doing That the Learner Should Own Next

• Who chooses the first task?
• Who identifies the weak link?
• Who decides whether to check?
• Who decides when to ask for help?
• Who decides what to revisit?
• Who recognises the finish condition?

The review should identify one responsibility ready for transfer.

Independence is not a final ceremony.

It is built through many small transfers of control.

47. The Review Should Ask Whether Feedback Changes the Next Attempt

Feedback has value only if it enters later performance.

The EEF’s current feedback evidence emphasises feedback that is linked to learning goals, gives specific next steps and is followed by opportunities to act.

So ask:

What correction from the last month can the learner now use without the tutor restating it?

If feedback remains trapped inside correction sessions, the route may need redesign.

48. The Review Should Ask Whether the Learner Can Monitor Themselves More Accurately

At the beginning, a student may say:

I’m bad at all of this.

Later:

I can do the algebra, but I am still slow at choosing the representation in mixed questions.

The second statement is itself progress.

The learner has a more accurate internal model of the problem.

EEF’s 2026 material on developing independent learners describes planning, monitoring and evaluation as processes that can become internal scaffolding over time.

The first review should look for that developing self-resolution.

49. What Formative-Assessment Evidence Contributes to the Review

The Institute of Education Sciences describes formative assessment as gathering, interpreting and using evidence about what and how students are learning so instruction can be adjusted while learning is still in progress.

That principle fits the first review closely.

The review is not a miniature final examination.

It is a decision point inside learning.

Gather enough evidence to decide whether the present instruction should continue, change or end.

50. What Research Does Not Give the Tutor

No evidence review can tell a private tutor that every learner should be reviewed after exactly four weeks using exactly the same metrics.

Research supports principles.

The tutor still needs professional judgement about:

• which evidence matters;
• how much change is educationally meaningful;
• which conditions should be compared;
• how much practice is enough before judging a route;
• when uncertainty is high enough to require another test.

Professional judgement becomes stronger when it is explicit enough to be challenged by later evidence.

51. Failure Mode: Continue Because the Student Has Already Paid

Payment history is not educational evidence.

The next lesson should exist because it has a useful job.

Commercial continuity and educational continuity are different questions.

52. Failure Mode: Advance Because the Student Is Bored

Boredom can be useful evidence.

It does not automatically mean the learner is ready for harder content.

The student may be bored because practice is too repetitive while transfer is still weak.

Advance the condition, not necessarily the chapter.

53. Failure Mode: Reclassify After Every Bad Lesson

A route that changes constantly cannot accumulate enough evidence to be judged.

One difficult day may reflect fatigue, unfamiliarity or ordinary variation.

Reclassify when the pattern, not the drama, changes.

54. Failure Mode: Stop Because Marks Have Not Moved Yet

Marks matter.

Some upstream improvements appear first as:

• better starts;
• fewer prompts;
• more accurate method selection;
• lower error recurrence;
• stronger delayed return;
• better correction use.

If those mechanisms are moving, the route may still deserve time.

The review should ask whether they are likely to convert into the target performance within the remaining horizon.

55. Failure Mode: Continue Because Marks Did Move

Better marks can hide a route that has already completed its job.

If the learner is independently stable, continuing identical support may add little value and begin to create dependence.

Success can be a reason to stop.

56. Failure Mode: The Review Becomes a Catalogue of Weaknesses

A thorough tutor can report twenty observations.

The review still needs one decision.

Compress:

What is the dominant bottleneck now, and what decision follows from that?

57. Failure Mode: The Tutor Reviews the Student but Not the Tutoring

Perhaps the learner is not the only thing that needs to change.

• Were prompts too frequent?
• Was practice too narrow?
• Was the tutor’s explanation too long?
• Was the class structure hiding the problem?
• Were school materials ignored?
• Was the route overloaded?

A mature review examines the tutoring intervention as part of the system.

58. Failure Mode: The Review Does Not Change Anything

The parent receives a long report.

The next lesson looks exactly like the previous one.

A review has value when it changes a decision:

• what to continue;
• what to remove;
• what support to fade;
• what condition to add;
• what tutor class to use;
• what route to stop.

If nothing changes, explain why continuation itself was the evidence-based decision.

59. The First Review Card

Field	Review question
Original hypothesis	What did we think was the main problem?
Original finish condition	What change were we trying to produce?
Observed change	What can the learner now do differently?
Support change	Has the smallest useful prompt reduced?
Delayed return	What survives later?
Transfer	What survives changed conditions?
School return	What has appeared in real school work?
Current bottleneck	What now blocks the greatest important progress?
Current tutor class	What tutoring function now dominates?
Current mode	Repair, Alignment or Frontier?
Decision	Continue, advance, reclassify or stop?
Next test	What evidence should the next cycle produce?

That is enough for a useful review.

The purpose is not to produce an impressive report.

It is to make the next decision better.

60. The Review Conversation With a Student

A student review can be short.

• At the start, what was hard?
• What is easier now?
• What can you now do without me?
• What still needs a cue?
• What happens when the question changes?
• What should we stop practising because it is already stable?
• What do you think the next challenge should be?

The tutor does not need to agree automatically.

The value is that the learner begins practising review itself.

61. The Review Conversation With a Parent

A parent review should answer:

• What was the original problem?
• What evidence repeated?
• What changed?
• What did not change?
• What can the learner now do with less support?
• What is the current bottleneck?
• What is the decision?
• Why?
• When will the next review happen?

For the wider family decision framework, see the Parents’ Guide to eduKate Sengkang.

62. The Review Should Produce a Smaller Plan, Not a Larger One

After a month, the tutor knows more.

That does not mean the plan should become more complicated.

Better evidence should often simplify the route.

We now know the bottleneck.

We now know what is already stable.

We now know what support can disappear.

We now know what should wait.

A good review reduces the number of things that need to be active at once.

63. The Next Route Should State a New Finish Condition

If the decision is continue, advance or reclassify, define the next finish condition.

For example:

Continue until Alicia can represent unfamiliar ratio relationships independently across two delayed mixed sets.

Or:

Advance to timed comprehension when Beatrice can identify the question relationship and select evidence accurately without tutor prompts across unseen passages.

Or:

Reclassify to Performance Coach work and review whether Ciara’s late-paper completion improves after two full-paper cycles using the exit-and-return protocol.

The next route should again be testable.

64. The Review Should State What Remains Unknown

Professional certainty should rise only where evidence earns it.

A tutor can say:

I am confident the original sign-handling weakness has improved. I am less certain whether the improvement survives long full-paper conditions, so that is the next thing we will test.

That uncertainty improves the next decision.

65. Twelve Questions Every Tutor Should Answer at the First Review

1. What was the original learner problem?
2. What change did we predict?
3. What change actually appeared?
4. What did not change?
5. What is the smallest useful prompt now?
6. What has survived a delay?
7. What has transferred to changed work?
8. What has returned to school or independent work?
9. What is now the dominant bottleneck?
10. What tutor class and tuition mode now fit?
11. Should we continue, advance, reclassify or stop?
12. What evidence will justify the next decision?

66. Twelve Questions Parents Can Ask at the First Review

1. What did you originally think the problem was?
2. What repeated evidence supported or changed that view?
3. What can my child now do independently?
4. What still needs help?
5. What support has been reduced?
6. What has transferred beyond tuition?
7. What is the current priority?
8. What are we deliberately not working on yet?
9. What is the decision for the next phase?
10. Why is that decision better than simply continuing the same work?
11. When will the route be reviewed again?
12. What evidence would make you change the plan?

67. Frequently Asked Questions

When should the first tuition review happen?

There is no universal date. A review should happen after enough repeated evidence has accumulated to judge the route fairly, or earlier if there is a major mismatch, safety concern, rapid improvement or major change in school demands. A month is a useful practical horizon for many weekly tutoring arrangements, but the evidence threshold matters more than the calendar.

What is the difference between a review and a progress report?

A progress report describes what happened. A review uses that evidence to decide whether the current route should continue, advance, change class or stop.

Should marks decide whether tuition continues?

Marks matter, especially when examination performance is the target. They should still be interpreted alongside the underlying mechanism, support conditions, delayed return, transfer and independent performance.

What if marks have not improved but the tutor sees progress?

Ask what progress means. Fewer prompts, better starts, stronger method selection, lower error recurrence and better transfer can be meaningful leading indicators. The tutor should explain how those changes are expected to convert into the target performance and by when.

What if marks improved but my child still needs a lot of help?

Test reduced-support and independent conditions. The improvement may be genuine but still heavily scaffolded. The review should distinguish learner growth from tutor-supported artifact improvement.

When should a tutor change strategy?

When repeated evidence shows that the present strategy is not producing the predicted change, when the learner has met the current finish condition, or when the dominant bottleneck has changed.

Is stopping tuition ever a positive outcome?

Yes. If the learner has achieved the intended independence and no longer needs the same level of support, reducing or stopping tuition can be evidence that the tutoring worked.

What if the student wants to stop but the parent wants to continue?

Return to the evidence and the purpose. Is the route producing useful change? Is the learner overloaded, mismatched or simply avoiding productive difficulty? The student’s view matters, but the decision should consider the full pattern rather than preference alone.

What if the tutor wants to continue but cannot explain why?

That is a problem. Continuation should have an explicit learner job, a finish condition and a reason another cycle is likely to be useful.

How much should the student be involved in the review?

Enough to understand what is changing, what still needs work, what responsibility is moving toward them and what the next route is trying to accomplish. The degree of participation should fit age and development.

What is the strongest sign that a route should advance?

The intended capability is independently stable enough that staying under the same support and difficulty produces little additional value, and the learner can handle a harder or more independent condition without the original weakness immediately returning.

What is the strongest sign that a route should stop?

Either the route has completed its job, or repeated fair tests show that it is not changing the intended learner capability and another intervention, structure or owner is more appropriate.

68. Evidence and Further Reading

• Education Endowment Foundation — Feedback and Developing Independent Learners, published 24 August 2026
• Education Endowment Foundation — Metacognition and Self-Regulation, published 24 August 2026
• Education Endowment Foundation — Feedback, Teaching and Learning Toolkit
• Education Endowment Foundation — One-to-One Tuition
• Education Endowment Foundation — Small-Group Tuition
• Institute of Education Sciences — Formative Assessment and Elementary School Student Academic Achievement: A Review of the Evidence
• eduKateSG — Tutor Classification Model
• eduKateSG — Tutorial Class Control Tower
• eduKateSG — The Three Modes of Tuition
• eduKate Sengkang — How Learning Diagnosis Works
• eduKate Sengkang — Finding the First Weak Link in Learning
• eduKate Sengkang — How We Know Learning Has Really Held
• The Tutor Handbook Vol No.0011 — The First Month

69. Final Compression

The first review asks one difficult question:

Does this route still deserve the learner’s next hour?

Not because the timetable says so.

Not because tuition has already started.

Not because the tutor likes the method.

Not because the parent has paid.

Because evidence says the next cycle has a useful job.

If the route is changing the intended capability and the finish condition has not been met, continue.

If the capability is stable enough for a harder or more independent condition, advance.

If the original explanation no longer fits, reclassify.

If the route has finished, failed, become dependent, become mismatched or crossed beyond tuition’s proper boundary, stop.

And after any decision, write down what evidence would make you change your mind next time.

A tutor becomes trustworthy not because every first plan is correct, but because the tutoring system is willing to notice when reality has earned a different next move.

That is the discipline of the first review.

That is Tutor Handbook Volume 0012.

Next in the series: The Tutor Handbook Vol No.0013 | The Parent Review — How a Tutor Explains Progress, Uncertainty and the Next Decision Without Turning the Child Into a Score.