The Tutor Handbook Vol No.0084 | The Dosage Differential

The Tutor Handbook · Volume 0084 · Series ID THB-0084

The Tutor Handbook: Complete series index.

The learner is improving.

Just not quickly enough.

The obvious response is to add more tuition.

Another weekday. A longer Saturday. More homework between lessons. A second tutor. A holiday intensive. A bigger package.

Sometimes that is exactly what the learner needs.

Sometimes it produces more of the same problem.

If Alicia has finally understood algebraic representation but meets it only once every seven days, insufficient exposure may be the bottleneck. More well-designed contact could help.

If Beatrice already attends twice a week but each session spends most of its time correcting completed answers rather than training evidence selection, another identical session increases dosage without improving targeting.

If Denise receives three sessions a week but does almost no independent work between them, the limiting variable may not be tutor contact at all. The learner may need a better bridge from taught performance to unsupported performance.

If Emily’s route is correct but school workload has made attendance irregular, the issue may be delivered dosage: the planned programme exists on paper but the learner is not receiving it consistently.

The Dosage Differential is the tutor’s decision about whether slow progress is best explained by too little exposure to a well-targeted route, enough exposure to a poorly targeted route, weak implementation of a good route, insufficient independent practice between contacts, or a different bottleneck that more sessions will not repair.

The direct answer is simple: do not increase tutoring merely because improvement is slower than hoped. First identify what additional tutoring would add that the present route is not already providing.

Research on tutoring gives good reason to take dosage seriously. Large reviews find positive average effects for tutoring, and several analyses report larger effects when tutoring is frequent and sustained. But frequency is not a magic variable. Programme design, tutor quality, targeting, attendance, student age, subject, group size and implementation all matter. “More” is therefore a decision to test, not a universal prescription.

1. What this volume owns

This volume owns a narrow tutoring decision: Should we increase the amount of tutoring, or would that simply increase exposure to the wrong mechanism?

It does not replace eduKateSG’s How Instructional Dosage Works | More Teaching Helps Only When the Additional Exposure Is Well Designed. That page owns the broader mechanism of instructional dosage.

It does not replace The Tutor Handbook Vol No.0038 | The Dose. Volume 0038 asks how much help to give inside an intervention before support begins replacing learning.

The Dosage Differential operates one level higher. It asks how much tutoring contact, across sessions and weeks, the route requires—and whether increasing that contact addresses the actual bottleneck.

2. “More tuition” contains several different changes

Parents often use dosage as though it were one variable.

It is not.

“More tuition” may mean:

more sessions per week;
longer sessions;
a longer programme across more weeks or months;
more practice between sessions;
more contact with the same tutor;
an additional tutor;
a short intensive period;
extra support only before assessments;
more feedback opportunities without increasing teaching time.

These changes do not produce the same learning conditions.

A second ninety-minute session may increase retrieval opportunities and shorten the gap between feedback cycles. Extending a single session from ninety minutes to three hours may instead increase fatigue while leaving the spacing pattern unchanged. Adding homework can increase independent practice without increasing tutor contact. Adding another tutor can increase contact while also introducing explanation inconsistency.

The first task is therefore to name the dosage dimension being changed.

3. Scheduled dosage is not delivered dosage

A programme can appear intensive on the timetable and be weak in reality.

Two sessions a week are scheduled.

One is repeatedly cancelled for school activities.

The learner arrives late to the other.

Homework intended to connect the two sessions is rarely completed.

On paper, the learner receives high dosage.

In operation, the route is fragmented.

Before prescribing more, distinguish planned dosage from delivered dosage. If attendance and between-session follow-through are unstable, increasing the schedule may simply enlarge the gap between the designed programme and the programme the learner actually receives.

4. Too little of the right tutoring has a recognisable pattern

The learner improves during sessions. Explanations make sense. Supported attempts become more accurate. The tutor can identify the correct weak link. But each new lesson begins with substantial rebuilding because too much of the route has decayed between contacts.

This is one pattern consistent with insufficient tutoring exposure.

Other signs may include:

the learner needs a useful feedback loop sooner than the current schedule provides;
the learner repeatedly forgets a newly repaired procedure before enough spaced retrieval has occurred;
school moves faster than the current contact frequency can support alignment;
a short intervention repeatedly ends before the repaired component has been reintegrated;
attendance is regular but opportunities to practise the target operation are genuinely sparse;
the route produces the expected receipt whenever it runs, but it does not run often enough for the gain to stabilise.

None of these automatically proves that another full lesson is necessary. The useful question is: what missing learning event would greater dosage provide?

5. Enough of the wrong tutoring has a different pattern

The learner attends regularly.

The sessions are long.

Homework is substantial.

Progress remains weak because the route repeatedly trains the wrong operation.

A student whose main difficulty is choosing an algebraic method receives more blocked worksheets where the chapter title already identifies the method.

A student whose main English weakness is evidence selection receives more model answers to memorise.

A student whose Science problem is causal explanation receives more keyword drills.

Each intervention may increase activity.

None necessarily trains the weak link.

When targeting is wrong, added dosage can strengthen the appearance of work while delaying the correct diagnosis.

6. Correct route, weak implementation is a third possibility

The plan may be sound while the execution is not.

The tutor intends to use mixed practice but repeatedly returns to blocked practice because it runs more smoothly.

The plan calls for the learner to attempt before receiving help, but answers are supplied quickly whenever hesitation appears.

Homework is designed as delayed retrieval but is completed immediately after the lesson with notes open.

The programme says support should fade, but the prompt remains permanently available.

Volume 0079, The Implementation Fidelity Check, owns this distinction in depth.

The dosage implication is important: do not add another session to compensate for a route that is not being implemented as designed. Fix implementation first, then observe whether dosage is still insufficient.

7. Between-session independence can be the missing dose

Not all useful dosage comes from a tutor.

A learner may need:

a ten-minute retrieval return two days later;
one fresh changed question;
a short self-explanation;
a small mixed set;
a planned correction check;
a school task used as transfer evidence.

If the route works in tuition but fades because nothing happens between lessons, the correct solution may be a better handover rather than another paid session.

This matters ethically and educationally. More tutor contact can help. It can also replace opportunities for the learner to carry the route independently if every useful learning event is scheduled with an adult present.

8. The first dosage question is “what is the bottleneck?”

Before adding sessions, ask which limitation currently dominates.

Exposure bottleneck: the right operation does not occur often enough.
Targeting bottleneck: the route trains the wrong thing.
Implementation bottleneck: the right plan is not being run faithfully enough.
Independence bottleneck: learning occurs with the tutor but is not carried between contacts.
School-alignment bottleneck: the route cannot keep pace with current external demands.
Capacity bottleneck: the learner has insufficient time or recovery to absorb more work.
Evidence bottleneck: progress may exist but the current measurement is too narrow or noisy to show it.

“More tuition” is only a direct answer to the first of these—and sometimes the fifth.

9. Frequency, session length and programme duration solve different problems

Frequency changes how often the tutor can teach, observe, correct and return to a target. Higher frequency can shorten feedback cycles and create more distributed contact.

Session length changes what can fit inside one encounter. A longer lesson may allow diagnosis, instruction, guided practice and an independent check to occur together. It may also introduce fatigue if length outruns useful concentration.

Programme duration changes how long the intervention can operate across time. Some repairs need enough weeks for practice, spacing, reintegration and delayed verification.

These dimensions can trade off. Three short meetings and one long meeting do not create the same learning conditions even if total minutes are equal.

Do not convert dosage into hours alone.

10. A learner can be under-dosed in one mechanism and over-loaded overall

Denise needs more distributed retrieval of one Additional Mathematics method.

She already has a full school week, two tuition subjects and heavy examination practice.

Adding another ninety-minute lesson may create enough exposure to the target but worsen sleep, recovery or time for other subjects.

The better dosage change may be a small, targeted retrieval return inside existing study time.

This is why The Learning Budget remains relevant. Dosage cannot be judged outside the learner’s total system.

11. More frequent tutoring can improve feedback speed without increasing total work

Suppose Alicia has one two-hour session each week.

She receives feedback on Monday, then attempts related school work through the week without another tutor check.

Changing the schedule to two one-hour sessions may leave total contact time unchanged while shortening the interval between attempt and professional review.

This may help if the active problem is repeated misapplication that needs earlier correction.

It may not help if travel, scheduling or context switching makes two sessions impractical. Dosage decisions are not abstract optimisations. They operate inside real family logistics.

12. Longer sessions can create false productivity

A three-hour session can cover more pages.

That is not the same as creating more durable learning.

If the final hour is dominated by fatigue, copying, passive review or tutor-led correction, the additional minutes may have low educational value.

Longer sessions make sense when the learning job genuinely needs a larger uninterrupted arc: perhaps a full diagnostic paper, extended writing cycle, experimental interpretation task or carefully staged performance rehearsal.

The question is not “Can we fill the time?” It is “What useful learning event requires this extra time?”

13. More sessions can create dependence

A learner who meets a tutor every day may receive rapid correction.

The same schedule can reduce opportunities to plan, attempt, recover and decide without adult presence.

This does not mean intensive tutoring is inherently dependency-producing. It means high dosage needs an independence design.

Ask:

Which decisions remain with the learner?
What happens between sessions without the tutor?
Does the learner attempt before receiving correction?
Is support being faded as capability grows?
Are there delayed checks outside the teaching context?
Does the programme contain an exit path?

More contact should increase capability, not merely increase access to rescue.

14. A dosage trial is better than a permanent escalation

If the evidence suggests under-dosage, test it.

Do not immediately commit the learner to a new permanent schedule.

A bounded trial might say:

For three weeks, split the current weekly contact into two shorter sessions so feedback arrives sooner. Keep total tutor time approximately constant. Preserve the same active target. Compare delayed independent performance and school transfer with the previous three-week baseline.

Or:

Add one twenty-minute independent retrieval block midweek without adding tutor contact. Review whether the Monday-to-Saturday decay pattern changes.

A dosage trial should alter the suspected limiting dimension while leaving enough of the route stable for interpretation.

15. Define the receipt before increasing dosage

What should greater dosage improve?

Do not settle for “more practice completed”.

Possible receipts include:

a repaired procedure survives a longer delay;
the learner needs less rebuilding at the next session;
a school task shows better transfer;
error recurrence falls;
independent study begins more reliably;
method selection becomes stable across changed problems;
the learner completes realistic sections without accuracy collapse.

If the tutor cannot name what additional dosage is supposed to change, the escalation is not yet well specified.

16. Define a cost signal too

17. Dosage should follow the route, not replace it

Frequency is a multiplier.

If the route is good, more well-designed exposure may amplify useful learning.

If the route is weak, greater frequency may amplify inefficiency.

This is why diagnosis comes before dosage. The learner needs the right operation before the system asks how often that operation should run.

18. A three-student class changes the dosage calculation

Three students can share one ninety-minute session and still experience different educational dosage.

Alicia receives two extended diagnostic interactions because her active route is uncertain.

Beatrice completes substantial independent work while the tutor observes another learner.

Ciara receives shorter but highly targeted feedback.

Clock time is shared.

Useful learning exposure is not necessarily equal minute by minute.

Do not judge dosage solely by class duration. Ask how often each learner performs the target operation, receives relevant observation, uses feedback and attempts independently.

19. A learner can need more frequency without needing more total tutor time

Scheduling architecture matters.

A learner who forgets a new procedure between weekly contacts may benefit from one short midweek check instead of a much longer weekend session.

A learner whose challenge is sustained examination performance may need the opposite: fewer fragmented meetings and an occasional longer integrated rehearsal.

Dosage should reflect the temporal structure of the target capability.

20. A learner can need less tutoring after improvement

Dosage is not a staircase that only moves upward.

If the learner can now plan, retrieve, practise and correct more independently, the correct dosage may fall.

This is not service withdrawal disguised as progress.

It is the intended educational direction when independent capability has genuinely grown.

The Release and The Independence Test own that later decision.

21. Composite case: Alicia needs more contact, not more content

This is an illustrative composite case.

Alicia has a well-defined weak link in algebraic method selection. During tuition she can explain the difference between factorisation, linear-equation solving and simultaneous equations. On fresh mixed work she still defaults to whichever method she practised most recently.

The tutor initially considers increasing worksheet volume inside the weekly lesson.

That would add more same-day practice.

Instead, the tutor notices the failure pattern: Alicia improves by the end of each session, then loses discrimination after six days without mixed selection practice.

The dosage hypothesis becomes temporal: the route needs a second contact with the operation before the weekly return.

Rather than adding a full lesson, the tutor assigns one short midweek independent mixed set and reviews only the method choices at the next meeting. If the decay pattern improves, the learner received more useful dosage without more tutor hours.

22. Composite case: Beatrice needs a different route, not another lesson

Beatrice attends English tuition twice a week.

Her comprehension marks remain flat.

The family considers adding a third lesson.

A review shows that most session time is spent going through model answers after Beatrice has completed passages. The tutor explains why answers are better, but Beatrice rarely has to choose evidence before receiving the model.

The bottleneck is not exposure to English tuition.

It is exposure to the target operation.

The route changes: shorter passage segments, evidence selection before answer construction, immediate changed examples and later independent return.

Only after that route has been implemented and observed would a dosage escalation be interpretable.

23. Composite case: Denise is receiving enough tuition but not enough independent performance

Denise has three Additional Mathematics contacts each week during an intensive period.

Her supported performance is strong.

Her independent school paper remains unstable.

Adding a fourth session would further increase supported contact.

The tutor instead asks what Denise does when no tutor is present. She rarely runs a full independent section because every practice block is either taught, corrected live or completed with reference material available.

The missing dose is independence.

One existing session is converted from guided practice to a representative unsupported performance check followed by later review. Total tutoring time does not rise. The evidence quality does.

24. Composite case: Emily’s schedule is the real constraint

Emily’s study route is carefully designed, but a temporary school project period causes repeated cancellations.

The parent sees missed sessions and proposes an additional weekly slot.

The tutor asks a different question: can Emily reliably attend the new slot, or will the programme simply contain more scheduled contact that is not delivered?

They temporarily shorten two sessions, protect attendance and preserve the most important target. Once the project period ends, the original schedule resumes.

Sometimes dosage is repaired by making delivery reliable rather than increasing the nominal total.

25. The Dosage Differential card

Current target: What capability is the route trying to change?
Current dosage: How many sessions, how long, across what programme duration?
Delivered dosage: What contact is actually occurring?
Target exposure: How often does the learner perform the actual weak-link operation?
Between-session work: What useful independent return happens without the tutor?
Implementation fidelity: Is the designed route really being run?
Receipt: What improves when the route runs correctly?
Decay pattern: What is lost between contacts?
School timing: Is the external curriculum moving faster than the route can support?
Capacity cost: What would added dosage displace?
Independence risk: Would more tutor contact replace a learning decision the learner should carry?
Smallest test: Can frequency, timing or independent practice be adjusted before adding a permanent session?

26. Parent conversation: “Would another lesson actually add a new learning event?”

Parents asking for more tuition are often responding rationally to visible risk.

Marks are low. Examinations are near. School moves quickly. The child appears uncertain.

Do not dismiss the request with “more tuition is not always better”. That slogan is too vague to help.

Instead explain the current dosage hypothesis:

She is responding well to the current algebra route, but the skill is decaying between weekly meetings. Before adding a full lesson, I want to test a short midweek retrieval return. If that reduces the rebuilding we see each Saturday, we will know the missing ingredient was frequency of practice rather than more explanation time.

Or:

He already receives enough contact. The issue is that our current practice is not forcing independent method choice. I would change the route before increasing the hours, then review whether dosage is still limiting progress.

Specificity turns a commercial-looking question into an educational decision.

27. Tutor conversation: protect yourself from dosage bias

Tutors are not neutral observers of dosage.

More sessions can appear professionally reassuring. The tutor has more control, more time to explain and more opportunities to correct.

That creates a bias toward interpreting uncertainty as a need for more contact.

Use a counter-question:

If I were not allowed to add another session, what would I change first?

If the answer immediately reveals a targeting, implementation or independence problem, dosage may not be the first move.

28. Research foundation: tutoring works on average, but programme features matter

Nickow, Oreopoulos and Quan’s systematic review and meta-analysis of experimental tutoring evidence, summarised by Stanford’s National Student Support Accelerator, synthesised 89 randomised evaluations and reported a pooled effect of roughly 0.29 standard deviations on learning outcomes. Their moderator analyses found larger average effects in some programme designs, including programmes meeting at least three days per week, while also showing substantial variation across contexts and designs.

This does not establish that an individual learner should attend tutoring three days a week. Meta-analytic moderators describe patterns across programmes, not deterministic prescriptions for one child. Programmes that meet more frequently may differ in other ways, and effects vary with age, subject, tutor type, implementation and study design.

The safe educational conclusion is narrower: dosage is a serious programme variable worth examining, not a magic number to copy.

29. Research foundation: high-quality tutoring is more than minutes

The U.S. Institute of Education Sciences describes high-quality tutoring as evidence-based support delivered with features such as trained tutors, frequent sessions, alignment and appropriate group size. Stanford’s Tutoring Quality Standards similarly treat dosage and ratio as important while placing them alongside tutor selection, tutor training and support, relationship quality, curriculum, formative assessment, data use and learning integration.

The implication is important for this volume: dosage sits inside a system of quality rather than substituting for that system.

Adding contact to a poorly targeted or weakly implemented route does not transform it into high-quality tutoring merely because the hour count increased.

30. Research foundation: implementation and monitoring remain part of the dosage decision

The Education Endowment Foundation’s current tutoring evidence summary reports positive average effects for tutoring while emphasising careful implementation, monitoring and evaluation. Its broader implementation guidance also treats implementation as a process rather than a one-off selection of an idea.

For a tutor, this means a dosage escalation should be reviewed like any other route change. Was the intended change delivered? Did the learner receive the new exposure? Did the expected receipt appear? Did a nearby cost increase? Did independence strengthen or weaken?

“We added another lesson” is an implementation fact.

It is not yet an improvement claim.

31. Failure mode: hours become the success metric

The programme reports eight hours of tutoring this month.

That may be useful operational information.

It does not tell us what changed in the learner.

Track the learning receipt beside the dosage.

32. Failure mode: low marks automatically trigger more sessions

A low score is evidence of a performance under conditions.

It does not tell us that insufficient tutor time caused the score.

Diagnose the loss before changing dosage.

33. Failure mode: the intensive rescue that never ends

An examination approaches, so temporary high-frequency tutoring begins.

The examination passes.

The intensive schedule remains because it became normal.

Temporary dosage should have an exit condition from the beginning.

34. Failure mode: replacing self-study with tutor time

The learner’s calendar becomes full of lessons.

There is little protected time left to retrieve, plan, attempt and correct without supervision.

The programme becomes highly supported and weakly independent.

High dosage should include unsupported receipts, not eliminate them.

35. Failure mode: changing dosage and route together

The learner moves from one to three sessions per week, receives new materials, changes tutor and begins a new study planner simultaneously.

Performance improves.

Which change mattered?

Volume 0065, The Concurrency Problem, explains why attribution becomes weak when many variables move together.

When possible, change dosage while preserving enough of the route to read the result.

36. Failure mode: copying programme averages into individual prescriptions

A research review finds stronger average results for programmes with frequent sessions.

The tutor concludes that every learner needs the same frequency.

That leap is not justified.

Volume 0071, The Applicability Check, exists for exactly this reason. Research can inform the hypothesis without dictating the individual route.

37. The ethical standard

Dosage decisions sit close to money.

That makes evidence discipline especially important.

A tutoring provider benefits commercially when families buy more sessions. A family may also feel safer when more time is purchased. Neither fact proves that the learner’s educational bottleneck is insufficient contact.

A professional tutor should be able to say, when the evidence supports it:

I do not think another lesson is the first thing to add. The current route needs a better independent return between sessions. Let us test that before increasing your cost and your child’s timetable load.

And when additional tutoring is genuinely justified:

The route is working when we run it, but one weekly contact is not enough to stabilise the repair before school moves on. I recommend a temporary second contact with a defined review point, and we will reduce it if the added exposure no longer earns its cost.

Both answers protect the learner.

38. Final principle

More tutoring is neither automatically better nor automatically excessive.

It is a dosage decision inside a learner-specific route.

Ask what additional contact will add.

Ask whether the route is correct first.

Ask whether the route is actually being delivered.

Ask what useful learning event is missing between sessions.

Ask what additional contact will displace.

Then test the smallest dosage change that can answer the question.

The right amount of tutoring is not the maximum a learner can tolerate. It is the minimum sufficient contact that allows the right learning route to take hold, travel beyond the lesson and become increasingly independent.

Small Group Tutorials

The Tutor Handbook Vol No.0084 | The Dosage Differential — How a Tutor Distinguishes Too Little Tutoring From the Wrong Tutoring Before Adding More Sessions