The Tutor Handbook Vol No.0064

The Tutor Handbook · Volume 0064 · Series ID THB-0064

The tutor has earned the right to change the route.

Then the learner presents five good reasons to change it.

Alicia needs more realistic timing in Mathematics.

Beatrice needs tighter evidence selection in comprehension.

Ciara is ready for less Science scaffolding.

Emily’s study plan needs a protected retrieval floor.

Faith should probably receive less prompting on familiar task forms.

Every proposed change is defensible.

That does not mean every proposed change should happen now.

The Change Queue is the tutor’s ordered set of justified but not-yet-simultaneous learning changes: what should move first, what should wait, what depends on something else, what is urgent, what is reversible, and what deserves to be rejected because the learner does not have enough capacity or evidence to run all good ideas at once.

The previous volume, The Tutor Handbook Vol No.0063 | The First Change, asked how a receiving tutor makes the first evidence-based modification without resetting the learner.

This volume owns the next systems problem:

What happens when more than one change is justified, but applying them together would make the learning system difficult to read, difficult to sustain, or difficult for the learner to carry?

Quick Read

A good change can be badly timed.
Several valid changes can compete for the same learner capacity.
The Change Queue separates deserves attention from deserves action now.
Queue order should come from leverage, dependency, urgency, reversibility, cost, observability and school timing—not tutor excitement.
The highest-leverage change is often the one that unlocks several downstream problems.
Some changes are prerequisites for later changes; sequence them accordingly.
Urgency can move a change forward, but urgency should not erase diagnostic discipline.
Low-reversibility changes deserve stronger evidence before entering the front of the queue.
A change that cannot be observed clearly may need a better measurement condition before implementation.
Do not run three repairs when one upstream repair may remove the other two.
Do not turn the queue into a backlog of permanent concern. Changes can be rejected, retired or rendered unnecessary.
The queue is not a to-do list. Each item needs a reason, a condition and a review rule.
Protect the learner from constant redesign.
Protect the tutor from mistaking busyness for responsiveness.
Protect parents from receiving five disconnected recommendations at once.
Protect school alignment when deadlines legitimately change the order.
Three-student tutorials can hold different queues for different learners.
Class 0–6 tutor functions see different kinds of queue items.
Repair, Alignment and Frontier modes use different ordering rules.
The learner should increasingly help read and order the queue.
The mature system has fewer active tutor-owned changes because the learner can prioritise more of the route independently.

1. What This Volume Owns

This volume owns sequencing when several learning changes are simultaneously plausible.

It does not ask whether a single proposed change is good. The Trial Run, Adoption Gate and Intervention Cost already own those decisions.

The Change Queue asks which good change should be allowed to consume the next unit of learner attention, tutor observation and programme stability.

2. A Valid Change Is Not Automatically a Current Change

Suppose a student has weak vocabulary, slow algebra, poor checking and an unstable study routine.

All four may deserve improvement.

The tutor still has to choose what enters the live route now.

If every weakness becomes an intervention, the learner receives a collection of correct observations and an unusable programme.

3. The Queue Exists Because Capacity Is Finite

Learners have finite attention, working memory, time, recovery and willingness to maintain routines.

Tutors also have finite observation bandwidth. If six things change, the tutor has less ability to tell which change produced which result.

Parents have finite coordination capacity. Schools have fixed deadlines. The timetable does not expand because the diagnosis became more accurate.

The queue is therefore not bureaucratic delay. It is a recognition that good interventions compete for scarce system capacity.

4. The Queue Is Not the Learning Budget

The Learning Budget owns the allocation of time, attention and practice across competing needs.

The Change Queue owns something narrower: the order in which modifications to the learning system are activated.

A learner may have enough weekly time for several tasks and still be unable to interpret several route changes at once.

5. The Queue Is Not a Wish List

“Improve vocabulary.”

“Do more full papers.”

“Work faster.”

“Be more independent.”

These are not queue items yet.

A queue item needs a current signal, a proposed change, a reason for priority, the evidence that would activate it, and the condition under which it can leave the queue without being implemented.

6. A Queue Item Can Expire

Beatrice’s tutor places “rebuild answer scope” in the queue.

Before it reaches the front, school work shows the problem has already stabilised.

The item should disappear.

A mature queue does not preserve old concerns merely because they were once true.

7. Order by Leverage

Leverage asks: if this change works, how much downstream difficulty might it remove?

A weak representation skill can damage several Mathematics topics. A small repair there may outperform separate topic drills.

A study-planning failure that causes repeated missed retrieval may affect several subjects. Repairing the planning mechanism can release more value than adding subject-specific homework.

High leverage does not mean “most important in theory.” It means “most likely to unlock the current system.”

8. Order by Dependency

Some changes are only useful after another condition is ready.

Do not add aggressive timing before execution is sufficiently stable.

Do not demand independent method selection if the learner does not yet understand the methods being selected between.

Do not fade a scaffold before the learner can perform the supported step reliably enough for removal to reveal learning rather than create chaos.

The queue should respect the dependencies of the learner’s actual route.

9. Order by Urgency

Urgency can legitimately reorder the queue.

A prelim paper next week may move examination timing ahead of a lower-risk vocabulary expansion.

A school assignment due tomorrow may temporarily outrank a longer-term study-system redesign.

But urgency should not become panic architecture. The tutor should still choose the smallest intervention that serves the deadline without destroying the longer route.

10. Order by Reversibility

A small change in question order is easy to reverse.

A major change in subject allocation, tuition load or learner responsibility is harder to reverse cleanly.

The harder a change is to reverse, the more evidence it should need before being allowed to jump the queue.

11. Order by Observability

Can the tutor actually tell whether the change worked?

“Try harder” has poor observability.

“Use one fresh mixed set without topic labels and compare method-selection accuracy with the baseline” has much better observability.

A poorly observable change should sometimes move backward in the queue until the tutor can define a better receipt.

12. Order by Cost

Two changes may promise similar gains.

One adds fifteen minutes twice a week.

The other adds forty-five minutes daily, increases parent supervision and competes with sleep.

Cost belongs inside priority, not as an afterthought.

A lighter intervention can deserve the front of the queue when it resolves the same uncertainty at lower system cost.

13. Order by Information Value

Sometimes the best next change is the one that teaches the tutor the most.

A short discriminating task may not directly improve the learner, but it may separate retrieval failure from concept failure and prevent weeks of wrong practice.

Information value can therefore move a small diagnostic action ahead of a larger intervention.

14. Order by School Timing

The learning system does not exist outside the school calendar.

A new chapter, weighted assessment, prelim period or subject transition can change which queued item is actionable now.

Alignment does not mean blindly following school order. It means respecting external constraints when they materially change opportunity cost.

15. Queue the Change, Not the Anxiety

Parents often notice many concerns at once.

“Her vocabulary is weak.”

“He is slow.”

“She does not revise enough.”

The tutor should convert concern into a bounded proposed change or leave it as an observation.

Anxiety should not acquire queue priority merely because it is emotionally loud.

16. Queue Items Need Activation Conditions

A useful queue item can read:

Add moderate timing only after method selection remains at or above the present accuracy floor across two fresh mixed sets.

This is better than “timing later.”

The queue now contains a condition that can become true.

17. Queue Items Need Exit Conditions

Not every queued change should eventually happen.

“Reopen fraction repair if the error recurs on two independent tasks.”

If the error never returns, the queued repair should expire.

Exit conditions stop the queue from becoming a museum of old worries.

18. Queue Items Need Owners

Who is responsible for observing the condition?

The tutor?

The learner?

The parent?

The school?

The queue is weak when everyone assumes someone else will notice the activation signal.

19. The Queue Should Stay Short

A tutor can identify twenty possible improvements.

The learner does not need to carry twenty future interventions.

Keep only the few changes that are both plausible and consequential enough to matter.

If an item is too vague, too distant or too weakly supported, keep it in ordinary notes rather than promoting it into the active change queue.

20. The Front of the Queue Should Be Explainable in One Sentence

“Why this change first?”

A strong answer sounds like:

Because the current error sits upstream of two other visible problems, the repair is low-cost and reversible, and next week’s school task depends on it.

If the tutor needs a long speech to defend priority, the queue may not yet be ordered clearly.

21. Do Not Confuse Severity With Priority

A serious-looking weakness may not be the first useful change.

The learner’s essays may look poor, but the high-leverage first change might be evidence selection rather than full composition practice.

A Mathematics paper may have many lost marks, but the first change may be interpretation of multi-step language rather than more calculation.

Priority belongs to mechanism, not visual drama.

22. Do Not Confuse Ease With Priority

Tutors naturally like changes that are easy to implement.

Print a worksheet.

Add a checklist.

Assign another paper.

Ease is useful, but it should not outrank diagnostic leverage.

23. Do Not Confuse Novelty With Priority

A new app, new framework or new resource can attract attention because it is new.

The queue should ask a less exciting question:

What current learner problem does this change solve better than the existing route?

24. The Queue Protects Stability Windows

A new idea appears during an active Stability Window.

Do not automatically insert it into the live route.

Queue it.

This allows the tutor to preserve the current experiment without forgetting the new observation.

25. The Queue Protects Rollback

When a route fails and rollback begins, queued innovations should usually wait.

Recovery is already a change-rich period.

Adding new interventions during rollback can make the recovery state unreadable.

26. The Queue Protects Cutover

Once a rebuilt route is being cut over into normal practice, avoid loading a second major redesign onto the same transition unless the learner requires it urgently.

Cutover should produce one stable operating state before the next structural change is activated.

27. The Queue Protects Tutor Handoffs

A Continuity Packet should distinguish:

what is currently active;
what is waiting in the change queue;
what activation condition belongs to each waiting item;
what concerns have already expired.

Otherwise the receiving tutor cannot tell future ideas from current instructions.

28. The Change Queue Card

Queue item: What proposed change is waiting?
Current signal: What evidence made the item worth tracking?
Priority reason: Leverage, dependency, urgency, cost, information value, school timing or another explicit reason?
Activation condition: What must become true before the change enters the live route?
Evidence required: What observation would justify activation?
Reversibility: How easy is it to undo?
Nearby cost: What could worsen?
Owner: Who watches the condition?
Expiry condition: What makes the item unnecessary?
Position: Why is it ahead of or behind other changes?

29. A Simple Queue Rule: One Structural Change, Several Small Adjustments

A practical default is to avoid running several structural changes together.

One major route change can coexist with small ordinary adjustments—clarifying an example, changing a question, moving a break—provided those adjustments do not alter the learner’s main job.

The rule is not mathematical. It is a discipline for keeping the route readable.

30. Sometimes Two Changes Belong Together

Not every paired change should be separated.

If support is reduced, a fresh independent task may be necessary to observe the effect.

If a new retrieval routine is added, the timetable may need a small allocation change so the routine has somewhere to occur.

Treat tightly coupled implementation steps as one intervention when separating them would make the change incoherent.

31. Bundles Need a Single Mechanism

A bundle is acceptable when its parts serve one mechanism and one receipt.

For example:

Remove the topic label, mix three familiar question families and ask the learner to name the method before solving.

Those are several operational moves serving one job: method selection.

By contrast, mixing question types, adding time pressure, removing prompts and increasing homework volume all at once creates several mechanisms and several possible causes.

32. Class 0 · Homework Helper Queue

Class 0 queue items are usually small operational routines: starting independently, recording tasks, packing materials, checking completion, or returning work.

Priority rule: change the routine that returns the most task-management responsibility to the learner without threatening actual school completion.

33. Class 1 · Explainer Queue

The Explainer’s queue should not become a list of topics to explain better.

First ask which topics still fail because structure is unclear. If explanation is no longer the bottleneck, move that item out of the Explainer queue and into practice, retrieval or transfer.

34. Class 2 · Drill Builder Queue

The Drill Builder often sees many possible practice upgrades: more spacing, more variation, less cueing, more speed, more mixing.

Priority rule: change the practice condition that targets the earliest reliability limit currently visible.

35. Class 3 · Diagnostic Tutor Queue

The Diagnostic Tutor may queue questions rather than interventions.

Which observation would separate two competing explanations?

The highest-value queue item may be the next discriminating probe.

36. Class 4 · Route Designer Queue

The Route Designer owns the clearest version of the queue because sequencing is the core job.

Queue order should integrate learner state, prerequisites, school timing and opportunity cost.

Do not let syllabus chronology replace dependency logic, and do not let dependency logic isolate the learner from the live school route.

37. Class 5 · Performance Coach Queue

Performance changes should be queued by the failure condition that most constrains real output.

Timing may wait behind method selection.

Full-paper endurance may wait behind section-level stability.

Checking routines may move forward when careless losses dominate otherwise stable work.

38. Class 6 · Learning Architect Queue

The Learning Architect can see many layers, which creates a special risk: too many plausible changes.

Class 6 discipline is therefore restraint.

The wider the system you can see, the more aggressively you must protect the learner from system redesign that is intellectually elegant but operationally excessive.

39. Repair Mode Queue

Repair Mode prioritises upstream, high-leverage weaknesses that currently block later performance.

Queue later refinements behind the weak link unless the school calendar creates a legitimate exception.

40. Alignment Mode Queue

Alignment Mode gives greater weight to current school demand.

However, alignment should not allow every new worksheet or test to seize the front of the queue.

Ask whether the external demand is high consequence, time-sensitive and genuinely related to the learner’s current participation.

41. Frontier Mode Queue

Frontier Mode has the largest temptation to accumulate exciting future work.

Queue extension by transfer value, not novelty.

Which next challenge builds a reusable way of thinking rather than merely moving the learner into harder content?

42. Three-Student Tutorials Need Three Queues

A shared classroom does not imply a shared change queue.

Alicia may have timing at the front.

Beatrice may have evidence discrimination.

Ciara may have scaffold fading.

The tutor can preserve common instruction while activating different next changes.

43. One Queue Item Can Become a Shared Mini-Lesson

If all three learners need the same operation—perhaps checking evidence against a question—the tutor can temporarily share the intervention.

But the item should enter each learner’s queue because the need is actually shared, not because group delivery is convenient.

44. Alicia · Mathematics: Timing Waits for Stable Selection

Alicia has two justified changes:

increase timing pressure;
introduce more varied representations.

Her current paper shows that method selection has only recently stabilised.

The tutor places moderate timing first because the examination clock is now the larger constraint and the representation change can wait one stability window.

The queue protects her from receiving every next-level demand in the same week.

45. Beatrice · English: Evidence Precision Before Vocabulary Expansion

Beatrice would benefit from richer vocabulary.

She also needs more precise evidence selection.

The comprehension evidence problem is currently costing direct marks and constraining explanation quality. Vocabulary expansion remains useful but moves behind it.

The queue says “not now” without saying “not important.”

46. Ciara · Science: Scaffold Fade Before Harder Content

Ciara is ready for more difficult Science material.

She is also still using visible model language before answering.

The tutor places scaffold fading first.

If precise explanations survive the fade, harder content becomes a cleaner next change because the learner will encounter it with more independent reasoning.

47. Denise · Additional Mathematics: Longer Sections Before Full Papers

Denise can eventually benefit from full-paper simulations.

Her queue still places longer representative sections first because this change increases performance load without simultaneously adding every full-paper demand.

Full papers activate after section-level accuracy and recovery remain stable.

48. Emily · Studying: Protect Retrieval Before Redesigning the Planner

Emily has two possible studying changes.

rebuild the weekly planner;
protect one small retrieval floor on high-load days.

The smaller change goes first because the current plan works except under overload.

Do not rebuild a system when one queue rule may solve the failure.

49. Faith · Independence: Remove One Prompt Before Adding Self-Marking

Faith is ready for less prompting and more self-checking.

The tutor queues prompt removal first.

Why?

Because self-checking evidence becomes more meaningful after the start of the task is genuinely hers.

Sequence improves interpretability.

50. Parent Communication: “These Are All Real, but We Are Not Doing All of Them Now”

Parents can mistake prioritisation for neglect.

A useful explanation is:

I agree that these three areas deserve work. I am not ignoring the other two. I am sequencing them because one is upstream, one is urgent for the next assessment, and changing all three now would make it hard to tell what is helping.

The parent now sees the queue as disciplined prioritisation rather than refusal.

51. Learner Communication: Show the Queue Without Showing a Burden

The learner does not need a ten-item improvement backlog.

Show the active change and perhaps one next item.

“This week we are working on choosing evidence precisely. Once that holds, we will decide whether vocabulary or timed writing deserves the next slot.”

The queue should create clarity, not anticipatory stress.

52. Learner Voice Can Reorder the Queue

A learner may report that a planned change is not the problem the tutor thinks it is.

Listen.

Then test.

Learner voice supplies information about effort, uncertainty, tool use and hidden support. It can change priority when the evidence supports the new interpretation.

53. The Queue Should Become More Learner-Owned Over Time

At first the tutor may identify and order most changes.

Later the learner can propose:

I think timing should wait. I am still choosing the wrong method when the diagram changes.

That is not merely preference.

It is emerging self-regulation: the learner can read the current state and argue for a better sequence.

54. Research Foundation: Implementation Is a Process, Not an Event

The Education Endowment Foundation’s A School’s Guide to Implementation emphasises that successful implementation depends on behaviours, context and process rather than simply selecting a promising idea.

The scale is different from one tutoring route, but the implementation principle transfers cleanly: a good idea still needs sequencing, fit, monitoring and sustained execution.

55. Research Foundation: Tutoring Needs Monitoring and Evaluation

The EEF’s Making a Difference with Effective Tutoring notes that tutoring is positively evidenced on average while also stressing the importance of careful implementation, monitoring and evaluation.

The Change Queue is one practical answer to that implementation problem. It helps the tutor avoid turning every observed need into simultaneous provision.

56. Research Foundation: Monitoring Should Change Teaching

The Australian Education Research Organisation’s Monitor Progress practice guide emphasises checking what students understand and can apply, identifying gaps and adjusting instruction, guidance or feedback where needed.

Monitoring does not require immediate action on every gap. It supplies the evidence from which queue order can be revised.

57. Research Foundation: Adaptive Teaching Is About Uncovering Learning

EEF’s 2026 discussion of adaptive teaching and feedback emphasises checking understanding and adapting teaching to what pupil responses reveal.

The queue adds a timing question: adaptation is strongest when the tutor not only notices what should change, but also knows which change should be activated now and which should wait.

58. Research Foundation: Student Agency Matters

The OECD Teaching Compass frames teachers as adaptive experts and co-designers of learning while placing student agency within the wider educational vision.

A mature Change Queue should therefore not remain permanently tutor-owned. The learner should increasingly contribute to the identification, ordering and review of future changes.

59. Common Queue Failure: Everything Is Priority One

If every weakness is urgent, the queue is not ordered.

Force the comparison.

Which change is more upstream?

Which has a nearer deadline?

Which is more reversible?

Which produces better information?

Which can wait without meaningful loss?

60. Common Queue Failure: The Oldest Item Always Goes First

A queue is not first-in, first-out by default.

New evidence can legitimately move a later item forward.

The queue exists to improve the learner, not to reward administrative age.

61. Common Queue Failure: The Tutor Keeps Adding Without Removing

Every review adds another concern.

Nothing expires.

The queue becomes a permanent accusation against the learner.

Review must delete as well as add. Stable gains, irrelevant concerns and obsolete hypotheses should leave the queue.

62. Common Queue Failure: The Queue Becomes a Hidden Curriculum

A tutor quietly stores twenty future changes and then releases them one by one without explaining the larger direction.

The learner experiences endless correction without understanding progress.

Use the queue to create sequence, not secrecy.

63. Common Queue Failure: No One Checks Whether a Waiting Item Is Still True

A queued item is a hypothesis about a future need.

Revalidate it before activation.

The learner may have changed while the item waited.

64. The Tutor’s Queue Audit

What changes are currently live?
What changes are waiting?
Why is each waiting item in the queue?
What is the activation condition?
What is the expiry condition?
Which item is highest leverage?
Which item is a prerequisite for another?
Which item is urgent because of school timing?
Which item is hard to reverse?
Which item is difficult to observe?
Which item has the highest cost?
Which item has the highest information value?
Can any two items be treated as one coherent intervention?
Am I queueing tutor preference rather than learner need?
Has any item become unnecessary?
Does the learner know the active change?
Does the learner need to know the whole queue?
What should become learner-owned next?

65. The Parent’s Queue Audit

Which problem is being changed now?
Which concerns are deliberately waiting?
Why is the current change first?
Is the order based on evidence or convenience?
What school deadline affects the order?
What would make a waiting issue move forward?
What would make it disappear?
Is the learner carrying too many new routines?
Are home requests adding unplanned changes to the tutor’s queue?
Is the tutor preserving enough stability to learn from each change?

66. The Learner’s Queue Audit

What am I changing now?
Why now?
What am I deliberately not changing yet?
What will show that this change is working?
What will I do when it becomes stable?
What next change do I think matters most?
What evidence supports my view?
Which change feels difficult because it is useful, and which feels difficult because the plan is badly designed?

67. What the Change Queue Is Not

It is not a punishment list.
It is not a catalogue of everything the learner is bad at.
It is not first-in, first-out administration.
It is not permission to delay urgent support.
It is not permission to ignore school deadlines.
It is not a substitute for diagnosis.
It is not a substitute for the Learning Budget.
It is not a substitute for the Stability Window.
It is not a promise that every queued change will happen.
It is an ordering discipline for justified future changes when the learner cannot responsibly run them all at once.

68. The Ethical Standard

A learner is not a project board.

They should not experience adult intelligence as an endless stream of redesign.

The more clearly a tutor can see possible improvements, the more responsibility the tutor has to protect the learner from unnecessary simultaneous change.

Good tutoring does not activate every good idea. It orders change so that the learner can carry it, the tutor can read it, the parent can understand it, and the next decision can be made from evidence rather than from the confusion created by the previous decision.

Evidence and Connected Reading

Final Compression

See the possible changes.

Do not activate them all.

Ask what is upstream.

Ask what is urgent.

Ask what depends on what.

Ask what is reversible.

Ask what the tutor can actually observe.

Ask what the learner can carry.

Ask what the school calendar changes.

Put one meaningful change at the front.

Give the others activation conditions.

Let stale items expire.

Let the learner eventually help order the queue.

The Change Queue turns “we should improve many things” into an executable learning route. It does not deny complexity. It gives complexity an order the learner can survive and the tutor can understand.

That is the Change Queue.

That is Tutor Handbook Volume 0064.

Small Group Tutorials

The Tutor Handbook Vol No.0064 | The Change Queue — How a Tutor Orders Several Valid Improvements Without Changing Everything at Once