A plain-English decoder for The Tutor Handbook.
The Tutor Handbook uses a lot of unusual words.
Gate. Check. Differential. Receipt. Route. Provenance. Fidelity. Fade. Branch Point. Frontier. Learning Claim. Support condition. Evidence freshness.
Then there are the tutor classes, the three tuition modes, diagnostic probes, transfer conditions, implementation terms and phrases such as “smallest useful intervention”.
If you have opened one of the Tutor Handbook articles and thought, What exactly does that word mean here?, this page is for you.
The terminology exists for a reason. Tutoring contains many decisions that look similar from the outside but are actually different.
A learner gets an answer wrong.
Do we explain again?
Give a hint?
Check an earlier prerequisite?
Ask another question?
Reduce the difficulty?
Increase the practice?
Wait?
Change the route?
Or investigate whether the learner actually understood the question?
Without precise language, all of those actions can collapse into the vague instruction: “Help the student.”
The Tutor Handbook tries to be more precise than that.
But precision becomes useless when the vocabulary itself becomes a barrier.
So this glossary translates the Handbook into ordinary English.
Some terms on this page are established language from education, assessment, learning science or implementation research. Others are eduKate Tutor Handbook control terms: labels created or adapted to make recurring tutoring decisions easier to recognise. A term appearing in the Handbook should therefore not automatically be read as a formal academic construct, qualification, medical concept or universally accepted educational category.
The point of the vocabulary is simpler.
It helps us ask:
What is happening? What do we think it means? What evidence do we have? What should the tutor do next? And how will we know whether that decision worked?
First: How to Decode a Tutor Handbook Title
Many Tutor Handbook titles become much easier to understand when you read them backwards.
Take:
The Implementation Feasibility Gate
The final word tells you the type of decision.
A Gate asks whether the conditions are sufficient to move forward.
“Feasibility” tells you what condition is being examined.
“Implementation” tells you where the problem lives.
So the title means:
Before we introduce this teaching practice more widely, can it actually work under ordinary conditions?
Now try:
The Support Provenance Check
A Check means inspect something before trusting the next conclusion.
“Provenance” means where something came from.
“Support” means the help that contributed to the learner’s performance.
So the question becomes:
Before treating this work as independent evidence, what help contributed to producing it?
Or:
The Missing-Response Differential
A Differential separates plausible explanations that would lead to different actions.
A “missing response” is the visible event.
So the real question is:
Was the question blank because the learner did not know, did not reach it, skipped it deliberately, misread it, could not access it, or ran out of time?
That is the basic grammar of much of the Handbook.
The strange-looking title is usually trying to compress a practical tutoring decision.
The Most Important Word: Route
Learning Route
A learning route is the tutor’s current working plan for getting from the learner’s present state to a more capable and independent one.
It is not simply a lesson plan.
A useful route connects what the learner is trying to learn, what the learner can presently do, the tutor’s current explanation of the problem, the teaching being used, the amount of support allowed, the evidence that will be collected next and the conditions that would make the tutor continue, change or end the plan.
A route is deliberately provisional.
It says:
“This is our best plan given what we currently know.”
It does not say:
“This is what we will continue doing regardless of what happens.”
That distinction runs through the whole Tutor Handbook.
Core Tutor Handbook Control Words
Gate
A gate answers:
Are the conditions good enough for us to make this move?
A gate normally has a consequence on both sides.
If the conditions are satisfied, proceed.
If they are not, hold, repair, investigate or choose another route.
Moving from supported work to independent work can have a gate. Increasing difficulty can have a gate. Introducing a new programme can have a gate. Fading coaching can have a gate.
A gate is not necessarily a test the learner passes or fails. Often it is a decision condition for the tutor.
Check
A check asks whether something we are relying upon is trustworthy enough for the next decision.
It is usually lighter than a gate.
We might check whether homework was independent, whether the task actually allowed the learner to demonstrate the target skill, whether two test results are comparable or whether the tutor’s own prompt accidentally created the error.
A check does not automatically demand major action.
It asks us to inspect before assuming.
Boundary
A boundary separates two legitimate activities that should not quietly become the same thing.
Support can become substitution.
Extension can become premature acceleration.
Tutor warmth can become inappropriate role expansion.
Preparation can become predetermining what a learner will do.
An accessibility support can be confused with answer-giving assistance.
The boundary tells us:
Both sides may be legitimate, but know when you have crossed from one job into another.
Differential
A differential separates several plausible causes when those causes would require different responses.
The Handbook borrows the reasoning pattern without using it as a medical diagnosis.
Suppose a learner suddenly performs badly.
Possible explanations might include missing knowledge, weak retrieval, unfamiliar wording, timing pressure, a prerequisite weakness, fatigue or a change in task difficulty.
The tutor does not merely list them.
The useful question is:
What small piece of evidence could distinguish between them?
Audit
An audit reopens a result that appears settled.
A correct answer can be audited.
A blank response can be audited.
A sudden improvement can be audited.
This does not mean distrusting learners.
It means recognising circumstances in which the visible result may hide something important.
If a learner who has struggled for several lessons suddenly answers perfectly after a model was left visible, it may be sensible to inspect how the answer was produced before declaring mastery.
Window
A window defines which evidence should count for a particular decision.
A very short window can overreact to one good or bad day.
A very long window can allow outdated evidence to dominate a learner who has genuinely changed.
An Evidence Freshness Window, Stability Window or similar Handbook mechanism is therefore asking:
Which period or sample gives us the fairest view of the present decision?
Protocol
A protocol is an agreed procedure for a recurring situation where unstructured judgement tends to drift.
It does not need to be complicated.
A handover protocol might specify what information should pass between tutors.
A disagreement protocol might specify what happens when a tutor and observer interpret a lesson differently.
A protocol provides structure without replacing professional judgement.
Trap
A trap names a predictable reasoning mistake.
A trap is not a learner problem.
It is often a warning to the adult making the decision.
Examples include mistaking several highly similar pieces of evidence for independent confirmation, overreacting to an unusually low score that naturally moves back towards the learner’s normal range, or allowing a strong overall score to hide an important weak component.
The word means:
Our intuition is particularly easy to fool here. Check before acting.
Receipt
A receipt is visible evidence that the intended change actually happened.
It is one of the most useful Tutor Handbook words.
Teaching happened.
Did learning change?
Feedback was given.
Did the next attempt change?
Tutor coaching happened.
Did live teaching change?
A programme was launched.
Did the intended practice actually appear under ordinary conditions?
The receipt is the evidence that comes back.
Agreement is not necessarily a receipt.
Enjoyment is not necessarily a receipt.
Completion is not necessarily a receipt.
The receipt must connect to the job we were trying to accomplish.
Probe
A probe is a small, deliberately chosen task or question used to reveal something.
A diagnostic probe is not simply another worksheet question.
Its value comes from what it distinguishes.
If two possible explanations predict different responses to one carefully chosen question, that question can be more informative than twenty routine exercises.
Threshold
A threshold is the amount or quality of evidence required before making a particular decision.
Not every decision needs the same threshold.
Changing one practice question is cheap and easily reversed.
Changing a learner’s whole programme, removing an important support or making a strong claim to a parent has greater consequences.
The Handbook therefore often connects the strength of evidence required to the cost of being wrong.
Branch Point
A Branch Point is the moment after enough diagnosis has occurred that the tutor must choose what kind of intervention comes next.
The learner may need explanation, practice, prerequisite repair, representation change, study-routine work, performance training or something else.
The Branch Point asks:
Given what we now know, which smallest intervention attacks the mechanism that actually matters?
Owner
An owner is the Handbook article responsible for one particular decision.
It does not mean a person owns the idea.
It means the website gives that decision one clear home.
If one article already answers “When should support fade?”, another article should not be created to answer exactly the same question under a slightly different title.
This makes a large Handbook easier to navigate.
Canonical Owner
The canonical owner is the main page responsible for a concept or decision across the eduKate ecosystem.
Other articles may mention the idea.
They should not quietly become competing versions of the same page.
Adjacent Owner
An adjacent owner covers the next genuinely different decision.
The Dose and the Fade are adjacent.
The Dose asks how much support to give.
The Fade asks how support should leave.
Closely related does not mean duplicate.
Collision
A collision occurs when two articles begin trying to own the same reader question.
Collisions matter because they make both the reader and the search system less certain about which page is authoritative.
The solution is not automatically deleting one page.
Often the correct repair is clarifying the boundary between two legitimate jobs.
Registry
The registry is the directory of Tutor Handbook volumes and their assigned decision jobs.
Think of it as the Handbook’s catalogue.
Router
A router helps a reader move from the problem they see to the article that owns the next decision.
A reader should not need to know that the page they require is Volume 0092.
They should be able to start with something ordinary such as:
“I think I know why the student is struggling, but how do I make sure my explanation isn’t wrong?”
The router can then lead towards a disconfirmation or diagnostic article.
Observation, Inference and Claims
Observation
An observation is what happened before we explain why it happened.
“Three denominator questions were wrong” is an observation.
“The learner has weak number sense” is an interpretation.
Keeping those separate prevents the tutor’s first theory from quietly becoming a fact.
Inference
An inference is the meaning we reasonably draw from evidence.
Inference is unavoidable in tutoring.
The important discipline is to know when we have moved from seeing something to explaining it.
Hypothesis
A hypothesis is the tutor’s current testable explanation.
“The learner may know the formula but fail to recognise when it applies” is a useful hypothesis because a tutor can design a task to test it.
“The learner is bad at Maths” is not useful because it is too broad, difficult to disconfirm and gives little guidance about the next teaching move.
Competing Explanation
A competing explanation is another plausible reason for the same visible event.
If two explanations require different interventions, the tutor should try to separate them before committing substantial teaching time.
Disconfirmation
Disconfirmation means actively looking for evidence that would show your preferred explanation is wrong.
This protects experienced tutors as much as new tutors.
Experience produces useful pattern recognition.
It can also produce premature certainty.
The question becomes:
If my current explanation were wrong, what might I observe?
Learning Claim
A Learning Claim is a statement about what a learner knows, can do, has improved, can do independently or is ready to do next.
The Handbook’s rule is simple:
The claim should never be stronger than the evidence.
“One familiar question was correct immediately after teaching” does not support the same claim as “the learner independently selected and executed the method a week later in mixed work”.
The stronger sentence must be earned by stronger evidence.
Causal Claim
A causal claim says one thing produced another.
“The new feedback routine caused the improvement” is much stronger than:
“Improvement followed the new feedback routine.”
Tutoring rarely provides laboratory control, so causal language should be used carefully.
Other things may also have changed.
Evidence Terms
Evidence
Evidence is information that helps a tutoring decision.
It might include learner work, verbal explanations, examination results, fresh questions, timing data, observations, marked scripts, homework conditions or implementation records.
More evidence is not always better evidence.
The useful question is:
Does this evidence reduce the uncertainty that matters for the next decision?
Discriminating Evidence
Discriminating evidence is evidence deliberately chosen because different explanations predict different results.
It is often more useful than simply gathering more data.
Fresh Evidence
Fresh evidence comes from a task or condition that is not merely a near-copy of what has just been taught or rehearsed.
Fresh does not mean completely unfamiliar.
It means fresh enough to reveal whether the learner can carry the relevant capability rather than reproduce the most recent route from memory.
Evidence Freshness
Evidence freshness describes how far the evidence has moved from the teaching, prompt, example or rehearsal that preceded it.
Immediate evidence can be useful.
It simply supports a narrower conclusion than delayed independent performance.
Provenance
Provenance means where something came from and what contributed to producing it.
For learner work, provenance can include help from tutors, parents, peers, notes, worked examples, answer keys or AI systems.
Knowing provenance does not make supported work worthless.
It tells us what claim that work can legitimately support.
Support Provenance
Support provenance asks specifically:
What help contributed to this performance?
A beautiful answer produced after heavy prompting may be excellent teaching evidence while remaining weak evidence of independence.
Those statements can both be true.
Triangulation
Triangulation means examining several differently conditioned sources of evidence to see whether they point towards the same interpretation.
A school test, a fresh tuition task and an independent explanation might provide useful triangulation.
Three nearly identical worksheets completed under the same conditions provide less independent confirmation.
Correlated Evidence
Evidence is correlated when apparently separate pieces of evidence share the same underlying weakness or condition.
Five worksheets generated from the same template may all flatter the same memorised method.
Counting them as five independent confirmations can create false certainty.
Comparability
Comparability asks whether two results can fairly be compared.
Were the tasks equally difficult?
Were support conditions similar?
Was the time allowance the same?
Was the marking standard stable?
Did the learner have the same legitimate access support?
A score difference means less when the conditions differ substantially.
Baseline
A baseline is the starting evidence against which later change is interpreted.
A baseline should be relevant to the capability being measured.
It is not necessarily one number.
Criterion
A criterion is the standard used to judge performance.
If the standard quietly changes, apparent improvement or decline may partly reflect the changed judgement rather than the learner.
Criterion Drift
Criterion drift occurs when the standard changes over time without being acknowledged.
A tutor may mark more generously after becoming familiar with a learner.
A programme may slowly redefine “implemented successfully”.
A criterion check protects comparisons from this drift.
Measurement Range
Measurement range asks whether a task is capable of showing meaningful differences in performance.
A test that is far too easy may hide improvement because almost everyone already scores near the top.
A task that is far too difficult may show only failure without revealing where the learner is changing.
The instrument must have room to see the change we care about.
Task Purity
Task purity asks whether a task mostly measures the capability we think it measures.
A difficult reading load can contaminate a Mathematics reasoning task.
Heavy handwriting demands can interfere with a task intended to assess idea generation.
Perfect purity is rarely possible.
The point is to notice important hidden demands before interpreting the result.
Measurement Contamination
Measurement contamination occurs when something outside the target capability materially contributes to the result.
Tutor prompting during an independence check is one example.
An answer-giving scaffold during a test of unaided method selection is another.
Opportunity to Demonstrate
Before concluding that a learner cannot do something, ask whether the task actually provided a fair opportunity to demonstrate it.
Missing evidence is not automatically evidence of missing capability.
Observation Reactivity
Observation reactivity means behaviour may change because somebody is watching.
A tutor being observed by a coach may teach differently from an ordinary Tuesday lesson.
A learner may behave differently when a parent sits beside the lesson.
Observation can still be valuable.
Its conditions simply belong in the interpretation.
Attrition
Attrition refers to people who begin a programme or sequence but disappear from later evidence.
If only the learners who remained are included in the final result, the result may look stronger than the experience of everyone who started.
Signal and Noise
Signal is the meaningful pattern we are trying to detect.
Noise is variation that does not represent the underlying change we care about.
Tutoring always contains some noise: question difficulty, tiredness, topic mix, ordinary good and bad days and random errors.
The goal is not to eliminate all noise.
It is to avoid redesigning the learning route every time noise moves.
Regression to the Mean
An unusually extreme result is often followed by one closer to a person’s usual range even when no powerful intervention has occurred.
That tendency is called regression to the mean.
If tutoring begins immediately after one unusually poor test, some rebound may have happened anyway.
This does not prove tuition had no effect.
It means the before-and-after story requires more evidence.
Learning and Performance Terms
Learning
In the Handbook, learning means a meaningful change in what the learner knows, understands, retrieves, represents, reasons with or can use.
Being taught something is not the same as learning it.
Completing something is not necessarily evidence that learning occurred.
Studying
Studying is what the learner does to build, organise, practise, retrieve or prepare knowledge and skills.
Studying can contribute to learning.
It is not identical to learning.
A learner can study for two hours without producing much durable change.
Teaching
Teaching is what an instructor does to create useful conditions for learning.
Successful teaching should be judged partly by what changes in the learner, not solely by how polished the explanation sounded.
Training
Training usually emphasises reliable performance through repeated, structured practice.
Training may support learning, fluency or performance.
The Handbook keeps it distinct because “understands why” and “can execute reliably under realistic conditions” are not always the same achievement.
Improvement
Improvement means a relevant change for the better.
It should still specify what improved.
Accuracy?
Understanding?
Speed?
Independence?
Transfer?
Method selection?
Performance under examination conditions?
“Improved” without a dimension can conceal too much.
Understanding
Understanding concerns the relationships, reasons, structures or meanings that organise knowledge.
A learner who understands can usually do more than reproduce a procedure.
But understanding alone does not guarantee fluent execution or successful performance under pressure.
Accuracy
Accuracy is producing the correct response or execution reliably.
Accuracy matters.
It is still only one dimension of performance.
Fluency
Fluency means reliable, increasingly efficient execution when efficient execution is genuinely useful.
Speed alone is not fluency.
A fast unreliable method is simply fast.
Retrieval
Retrieval means bringing knowledge back from memory rather than only recognising it when it is visible.
Retrieval matters because learners frequently need knowledge when the example, notes or answer choices are absent.
Retention
Retention means learning remains available after time has passed.
Immediate success and retained learning are different claims.
Transfer
Transfer means using learning when something relevant changes.
The wording may change.
The representation may change.
The context may change.
The method label may disappear.
The tutor may change.
The task may become mixed.
The learner may move from tuition to school.
Transfer is therefore not simply “doing a harder question”.
It asks whether the underlying capability travels.
Generalisation
Generalisation is the ability to use an underlying idea across appropriate examples, contexts or situations rather than only in the exact form in which it was learned.
It overlaps with transfer, although the Handbook often uses transfer for the practical decision of testing whether learning survives changed conditions.
Method Selection
Method selection means deciding which method to use when the task does not announce the answer.
A learner may execute three methods perfectly when worksheets are separated by chapter and still struggle when those methods are mixed.
That is a selection problem, not necessarily an execution problem.
Performance Under Conditions
This means producing the capability under the conditions that eventually matter.
Time pressure.
A full examination paper.
Mixed topics.
Independent work.
A particular response format.
Teaching and verification should eventually include the relevant real condition without introducing every difficulty at once.
Support and Independence Terms
Support
Support is help that makes productive learning or performance possible.
Support can be instructional, organisational, environmental or technological.
The central question is not whether support exists.
It is what job the support is performing.
Scaffold
A scaffold is temporary guidance or a resource that helps a learner perform something they cannot yet manage adequately alone.
Examples include worked examples, prompts, diagrams, sentence frames, checklists and partially completed problems.
Good scaffolding has a direction towards greater learner ownership.
The Australian Education Research Organisation describes scaffolding as support used to help students achieve learning objectives and distinguishes planned scaffolding from support introduced responsively during lessons. Its current practice guide also emphasises gradually removing scaffolds as proficiency develops. AERO: Scaffold practice.
Planned Scaffold
A planned scaffold is designed before the lesson because the tutor reasonably expects it will be useful.
Contingent Scaffold
A contingent scaffold is introduced in response to what the tutor observes during the lesson.
The distinction matters because a tutor should be prepared without pre-deciding exactly how the learner will fail.
Prompt
A prompt is a cue intended to help the learner initiate, continue or correct a process.
Prompts vary enormously in how much thinking they provide.
“What should you check first?” and “Use the quadratic formula” are both prompts, but they carry very different amounts of the problem.
Dose
Dose is the Handbook’s metaphor for how much support is being given inside an intervention.
It does not imply medicine or diagnosis.
One cue and six cues are different doses.
One example and four examples are different doses.
A temporary checklist and a tutor managing every step are different doses.
The Dose asks:
How much help is enough to make progress without allowing the help to take over the learner’s job?
See also The Tutor Handbook Vol No.0038 | The Dose.
Dosage
The Handbook uses dosage more broadly for how much tutoring or instructional exposure occurs across sessions and time.
This is different from Dose.
Dose asks how much help is given inside the interaction.
Dosage can concern session frequency, duration or amount of tutoring.
The National Student Support Accelerator’s tutoring-program toolkit likewise treats dosage as a distinct programme-design dimension. National Student Support Accelerator: Toolkit for Tutoring Programs.
Fade
A Fade is the controlled removal of a support as the learner becomes able to carry more of the process.
Fading does not mean suddenly saying:
“Do it yourself.”
A good fade transfers meaningful pieces of responsibility and watches what happens.
The Handbook’s key idea is that support should leave in a way that reveals ownership rather than causing avoidable collapse.
See also The Tutor Handbook Vol No.0039 | The Fade.
Independence
Independence means the learner can perform the relevant job without the support that previously carried it.
It does not mean never asking for help.
A sophisticated independent learner may recognise a genuine limit, request appropriate help, use that help and resume responsibility.
Self-Regulation
Self-regulation means increasingly managing one’s own learning processes: planning, monitoring, checking, adjusting and recovering when something goes wrong.
The tutor’s long-term job is not to become the learner’s permanent self-regulation system.
Learner Agency
Learner agency means the learner increasingly participates meaningfully in decisions about goals, strategies, monitoring and next actions.
Agency is not simply letting the learner choose everything.
Useful agency operates within appropriate educational boundaries.
Accessibility Support
An accessibility support helps a learner gain fair access to the task.
It should not automatically be removed in order to “prove independence”.
A support that is legitimate in the learner’s real environment may remain legitimate during verification.
The important distinction is whether the support provides access or performs the target skill on the learner’s behalf.
Repair, Verification and Release
Repair
Repair is teaching intended to rebuild a weak or unstable capability.
Repair usually allows appropriate support.
That matters because repair and verification serve different purposes.
Verification
Verification asks whether the capability still appears when the relevant support or condition changes.
A learner may need support during repair and less support during verification.
Using verification conditions too early can make teaching unnecessarily harsh.
Using repair conditions too late can make success look more independent than it is.
Reattempt
A reattempt is another attempt after teaching, feedback or correction.
Its meaning depends on conditions.
Repeating the exact question immediately after seeing the solution provides different evidence from attempting a fresh version later without the model.
Changed Condition
A changed condition deliberately alters one relevant feature to test what the learner now carries.
That feature could be time, wording, representation, support, context, task ordering or delay.
Changing everything at once makes failure difficult to interpret.
Maintenance
Maintenance keeps a confirmed gain alive without rebuilding it from the beginning.
Maintenance should normally be lighter than repair.
Reopening
Reopening means a previously repaired weakness has become consequential again and needs renewed attention.
Reintegration
Reintegration puts a repaired component back into the larger performance from which it was isolated.
A learner may repair one algebraic step successfully in narrow practice.
Reintegration asks whether that repaired step survives inside full problems again.
Release
Release means the learner or route no longer needs the current level of tutoring support.
Release is a positive tutoring outcome.
Good tuition should contain some route towards needing less of the tutor for the capability that has been built.
Escalation
Escalation means the present tutoring response is no longer sufficient and a larger, different or more specialised action is justified.
Escalation should have a reason.
It should not simply mean “do more because the current thing did not work”.
The Tutor Classification Model
The Tutor Classification Model describes tutoring functions.
It is not an accreditation system.
It is not a permanent ranking of human beings.
A tutor may perform several functions within one relationship.
The useful question is:
Which tutoring job is required now?
The early Handbook volumes establish the following Class 0–6 structure.
Class 0 — Homework Helper
The dominant job is making assigned work operable while keeping as much thinking as possible with the learner.
Useful when organisation, initiation, task interpretation, checking routines or completion are the main problems.
Class 1 — Explainer
The main job is making a difficult concept, relationship or method understandable.
The Explainer should clarify without becoming the permanent source of every answer.
Class 2 — Drill Builder
The learner broadly understands the idea.
The main need is stabilising execution through well-designed practice.
Drill should build reliability rather than mindless repetition.
Class 3 — Diagnostic Tutor
The visible mistake is not enough.
The tutor investigates where the consequential weak link actually lies and separates plausible causes before committing to more teaching.
Class 4 — Route Designer
Several learning needs compete for time.
The Route Designer decides what should happen first, what can wait, what should be maintained and how learning should unfold across time.
Class 5 — Performance Coach
The capability substantially exists, but the learner cannot yet convert it into reliable output under the conditions that matter.
Timing, method selection, full-paper control, pressure, recovery and performance routines may dominate.
Class 6 — Learning Architect
Several interacting parts of the learning system need coordination.
The Learning Architect can move between lower functions, but Class 6 does not mean “do everything”.
The defining skill is integration and restraint: understanding which part of the system should change and which parts should remain deliberately stable.
The highest class is therefore not automatically the best tutoring choice.
If a learner simply needs one idea explained, a clean Class 1 intervention may be better than unnecessary Class 6 complexity.
The Three Tuition Modes
The Handbook also uses three broad tuition modes.
These describe what the tuition is trying to accomplish, not the learner’s permanent ability.
A learner can be in different modes for different domains.
Repair Mode
Repair means rebuilding a consequential earlier weakness because it is interfering with current or future learning.
The tutor asks:
What earlier instability is creating downstream cost?
Alignment Mode
Alignment means keeping the learner effectively connected to current school or curriculum demands.
The learner may not need reconstruction of everything.
The problem is often staying coherent with what must be understood, completed or performed now while still protecting deeper foundations.
Frontier Mode
Frontier means the learner is sufficiently stable for depth, extension, abstraction, transfer, synthesis or more demanding performance to become a major priority.
Frontier does not simply mean “give harder worksheets”.
The challenge should increase what the learner can think, choose, explain or transfer.
Series Frontier
The Handbook also occasionally uses frontier to mean the newest end of the publishing series.
That is completely different from Frontier Mode.
“Live through Volume 0210” describes the series frontier.
“Frontier Mode” describes a tutoring purpose.
Context tells you which meaning is intended.
Diagnosis Terms
Symptom
A symptom in Handbook language is the visible problem.
A wrong answer.
A blank question.
A sudden mark drop.
Very slow work.
Repeated forgetting.
The word is descriptive here, not medical.
The important principle is that the visible symptom is not automatically the cause.
Weak Link
A weak link is an unstable component whose failure meaningfully limits later performance.
The useful weak link is not always the earliest thing a learner ever found difficult.
It is the weakness currently exerting consequential downstream cost.
Bottleneck
A bottleneck is the part of the system currently limiting progress.
Improving other things may make little difference until the bottleneck changes.
The bottleneck can move.
That is why the learner should not be permanently labelled by yesterday’s bottleneck.
Prerequisite
A prerequisite is knowledge or capability that needs to be sufficiently available before another capability can operate well.
Prerequisite repair should travel backwards only as far as necessary.
Otherwise tutoring can become an endless excavation of old material.
First Weak Link
The first weak link is the earliest consequential breakdown in the chain relevant to the present task.
It is not necessarily the first error written on the page.
Practice Terms
Worked Example
A worked example shows a completed route through a problem or task.
Worked examples can reduce unnecessary load while a learner is building a new process.
They eventually create an important question:
When should the learner stop studying the route and begin generating more of it?
Contrast Case
A contrast case places examples side by side so the learner can notice a decisive difference.
Good contrast cases change something that matters.
Poor contrast cases simply create more material to inspect.
Blocked Practice
Blocked practice groups similar problems together.
It can be useful while stabilising a method.
But blocked practice can hide a future method-selection problem because the page itself tells the learner which method is required.
Mixed Practice
Mixed practice places different problem types together.
The learner must decide what kind of problem is present and which method fits.
It therefore tests something different from repeated execution of one announced method.
Spaced Practice
Spaced practice distributes practice across time rather than concentrating everything into one sitting.
Spacing can support longer-term retention and creates opportunities to see what remains available after the immediate teaching context has faded. AERO’s current practice guidance combines spacing and varied practice as ways to strengthen retention, fluent recall and adaptable use. AERO: Vary practice.
Variation
Variation changes selected features of practice so the learner learns what is essential and what can change.
Variation becomes especially useful when preparing for transfer.
Group Tuition Terms
Orchestration
Orchestration is the tutor’s management of several learners, tasks, turns, observations and feedback needs inside one coherent lesson.
Good small-group tuition is not simply individual tutoring performed three times simultaneously.
The group has shared moments and individual routes.
The tutor must know when those should converge and when they should separate.
Individual Accountability
Individual accountability means each learner eventually needs to produce evidence of their own understanding or performance.
A strong group answer cannot automatically be treated as evidence that every individual learner can do the work.
Group Leakage
Group leakage occurs when information intended for one learner gives away too much of another learner’s task or answer.
One student’s response can unintentionally become another student’s prompt.
Shared Error
A shared error is an error sufficiently common across the group that shared reteaching may be efficient.
Tutors should still verify that the learners are failing for substantially the same reason.
Three identical wrong answers can arise from three different mechanisms.
Case Mix
Case mix refers to the combination of learner needs a tutor or programme is managing.
A group may become difficult not because any individual learner is inappropriate, but because the combined needs demand more simultaneous control than the lesson format can support well.
Feedback and Coaching Terms
Feedback
Feedback is information intended to help improve the next relevant performance.
Its value is not established merely because it was delivered.
Feedback Uptake
Feedback uptake asks whether the learner actually uses feedback in later action.
The feedback was read.
Did anything change?
That change is the more important evidence.
Coaching
In the Tutor Handbook, coaching usually refers to professional development that helps a tutor improve live teaching.
It should not be confused with learner performance coaching, which is a Class 5 tutoring function.
Context matters.
Coaching Receipt
A Coaching Receipt is evidence that professional feedback changed what the tutor does in actual teaching.
Explaining the new technique in a meeting is not enough.
The professional-learning loop is completed when the intended move appears reliably in practice.
Live Practice
Live practice is real teaching under ordinary working conditions rather than explanation, rehearsal or simulation.
A tutor can perform beautifully in a workshop exercise and still struggle to use the same judgement while managing real learners.
Programme and Implementation Terms
As the Tutor Handbook expanded from individual tutoring into tutor coaching and programme quality, it adopted terminology that is common in implementation work.
The Education Endowment Foundation’s current implementation guidance emphasises a structured implementation process, contextual factors and the day-to-day enactment of evidence-informed approaches. EEF: A School’s Guide to Implementation.
Implementation
Implementation is the work of making an intended practice actually happen under real conditions.
A good idea on paper is not yet an implemented practice.
Active Ingredient
An active ingredient is a component believed to be important to how a practice produces its intended effect.
Implementation becomes difficult when nobody knows which parts must remain stable and which parts can sensibly adapt.
The active-ingredient idea helps distinguish where implementation should be tighter and where local flexibility may be appropriate.
Fidelity
Fidelity asks whether the important parts of the practice were actually carried out as intended.
If a programme was barely implemented, a poor result cannot automatically establish that the underlying approach failed.
At the same time, fidelity should not become an excuse for forcing every superficial detail to remain unchanged.
Adaptation
Adaptation means changing a practice to fit the context.
Useful adaptation preserves what matters while changing what can legitimately vary.
The key question is:
Did we adapt the delivery, or did we remove the mechanism that made the practice worth using?
Feasibility
Feasibility asks:
Can this practice realistically be carried out here?
Tutor time, materials, timetables, class sizes, workload, systems and learner variation all matter.
A practice can be theoretically strong and operationally impossible.
Acceptability
Acceptability asks how people experience and respond to the practice.
Do tutors find it workable and professionally credible?
Do learners understand it?
Do families tolerate the burden?
Acceptability matters because practices that people cannot reasonably live with tend not to survive.
It is not the same as effectiveness.
A popular intervention can be ineffective.
An initially unpopular change can still be educationally valuable.
Reach
Reach asks whether the intended people actually received the practice.
A programme can show good fidelity among the tutors who use it while reaching only a fraction of the learners it was designed to serve.
Sustainability
Sustainability asks whether the practice can remain alive after the launch energy, temporary staffing or enthusiastic champion disappears.
If a programme works only while extraordinary effort is being supplied, it may not yet be sustainable.
Readiness
Readiness asks whether enough of the necessary conditions exist before implementation begins.
Are the problem and purpose understood?
Are materials ready?
Is there sufficient time?
Can tutors be trained and supported?
Can the programme observe what is happening?
Launching before readiness can make a sound idea appear weak.
Rollout
A rollout is the introduction of a practice to the people or settings expected to use it.
Rollout is not the same as implementation success.
Sending the materials is not the same as changing practice.
Scale
Scale means expanding a practice to more people, groups, sites or contexts.
Scaling changes the implementation problem.
A practice that works with one highly experienced tutor may need different systems when twenty tutors must run it consistently.
Implementation Outcome
An implementation outcome tells us something about how the practice is being installed or sustained.
Fidelity, reach, feasibility, acceptability and sustainability are examples.
They should not automatically be averaged into a single “implementation score” because each answers a different question.
AI and Tool Terms
AI-Assisted Work
AI-assisted work is learner or tutor work to which an AI system materially contributed.
The important issue is not simply whether AI was present.
It is what job the AI performed.
Generating examples is different from generating the learner’s reasoning.
Checking grammar is different from constructing the argument.
AI Material Verification
AI Material Verification means checking generated material before trusting it for teaching.
AI output can be fluent and still contain errors, invented references, unsuitable difficulty, curriculum mismatches or poor examples.
Responsibility remains with the human using the material.
Human-Supported AI Engagement
This describes AI use in which a human tutor remains responsible for purpose, interpretation, boundaries and verification.
The tool can assist the tutoring system.
It does not automatically become the professional decision-maker.
Learner Data
Learner data is information about the learner created or collected during tutoring.
Examples include work, scores, observations, recordings, attendance, behaviour notes and AI interaction records.
The educational usefulness of retaining information should be balanced against privacy, relevance and appropriate access.
A Few Easily Confused Pairs
Support versus Substitution
Support helps the learner perform the target process.
Substitution occurs when the helper performs the crucial part instead.
The line is sometimes subtle.
That is why the Handbook pays so much attention to support conditions.
Teaching versus Evidence
A scaffold can be excellent for teaching while making a task poor evidence of independence.
That is not a contradiction.
It simply means the task’s purpose matters.
Difficulty versus Diagnostic Value
A harder question is not automatically a better diagnostic question.
The best diagnostic question is the one that helps distinguish the explanations that matter.
Engagement versus Learning
A learner can be highly engaged without learning much.
A learner can also experience demanding learning without appearing constantly enthusiastic.
Engagement matters.
It is not a substitute for learning evidence.
Confidence versus Capability
Confidence is a learner’s belief or feeling about performance.
Capability is what the learner can actually do under defined conditions.
The two interact, but they should not be treated as the same variable.
Practice versus Verification
Practice is designed to improve the capability.
Verification is designed to learn what the learner can now carry.
The allowed support may therefore be different.
Extension versus Acceleration
Extension deepens, broadens or transfers capability.
Acceleration moves into material normally encountered later.
They can overlap.
They are not identical.
Tutor Consistency versus Tutor Rigidity
Consistency means important principles remain dependable.
Rigidity means continuing the same action despite evidence that the route should change.
The Handbook aims for the former, not the latter.
The Handbook’s Main Decision Loop
Most Tutor Handbook terminology eventually returns to one practical sequence:
- Observe what happened.
- Define the capability that matters.
- Identify the uncertainty most likely to change the next action.
- Collect enough discriminating evidence.
- Separate plausible explanations.
- Choose the smallest useful intervention.
- Teach.
- Transfer responsibility.
- Verify on fresh or changed-condition evidence.
- Review the route.
- Then continue, advance, hold, repair, change, release or escalate.
That loop matters more than memorising every term.
The vocabulary exists to make the decisions inside the loop easier to see.
For Parents: The Five Questions That Translate Most of the Handbook
A parent does not need to learn two hundred article titles.
Most of the Tutor Handbook can be translated into five ordinary questions.
What are we trying to improve now?
What makes you think that is the real problem?
What help is my child currently receiving?
What can my child now do without that help?
What evidence would make us continue, change or end the present plan?
Those questions contain most of the architecture.
For Learners: An Even Shorter Translation
A learner can translate the Handbook into:
What is my job?
What help can I use?
What part should I try myself?
How will I know I am getting better?
What do I do if it stops working?
That is a learning route in ordinary language.
For Tutors: The Point of All This Terminology
The vocabulary should make tutoring clearer.
Never let it make tutoring look more sophisticated than it is.
Calling something a “gate” does not make the decision good.
Calling something a “differential” does not make the diagnosis correct.
Calling something a “receipt” does not make weak evidence strong.
Calling something “Class 6 architecture” does not justify unnecessary complexity.
The terminology earns its place only when it helps the tutor make a better decision.
If the plain-English question is clearer than the technical term, use the plain-English question.
The learner does not need jargon.
The learner needs good tutoring.
The parent does not need a performance of expertise.
The parent needs to understand what is happening and why.
The tutor does not need two hundred labels in memory.
The tutor needs to recognise the next consequential decision.
That is why the Tutor Handbook has a glossary.
Not to make the system sound complicated.
To make a complicated human activity easier to think about.
Terminology and Evidence Note
The Tutor Handbook combines several kinds of vocabulary.
Some terminology is ordinary educational language. Some comes from learning, assessment and implementation research. Some is a specific eduKate operating label created to make tutoring decisions easier to distinguish.
This glossary therefore uses the following rule:
When the wider education field already has an established concept, preserve its established meaning. When the Tutor Handbook introduces a control label, explain it as a Handbook convention rather than pretending it is universal academic terminology.
Current external reference points include the Australian Education Research Organisation’s practice guidance on scaffolding and varied practice; the Education Endowment Foundation’s implementation guidance; the National Student Support Accelerator’s tutoring-program toolkit; and the US What Works Clearinghouse practice-guide framework. These sources support many of the broader instructional and implementation concepts while the Tutor Handbook remains responsible for its own internal decision language.
- Australian Education Research Organisation — Scaffold practice
- Australian Education Research Organisation — Vary practice
- Education Endowment Foundation — A School’s Guide to Implementation
- National Student Support Accelerator — Toolkit for Tutoring Programs
- What Works Clearinghouse — Practice Guides
The glossary should continue to grow when a genuinely recurring term needs translation.
It should not grow merely because another impressive-sounding phrase can be invented.
The test is simple: does defining the term make the Tutor Handbook easier for a real tutor, learner or parent to use?
If yes, it belongs here.
If not, plain English wins.
Return to The Tutor Handbook.
