The Tutor Handbook Vol No.0068 | The Opportunity Check

The Tutor Handbook · Volume 0068 · Series ID THB-0068

The Tutor Handbook: Complete series index.

A tutor asks a student to explain a Science mechanism. The student gives three hesitant sentences and stops.

The easy conclusion is that the learner does not understand the mechanism well enough.

That may be true. It may also be wrong.

Perhaps the learner can build the mechanism accurately with a diagram but cannot yet compress it into fluent speech. Perhaps the question silently depends on vocabulary the learner does not know. Perhaps the tutor interrupted after every clause, so the learner never had an uninterrupted opportunity to assemble the answer. Perhaps the learner was asked to perform a skill that has never actually been taught. Perhaps an access support normally used at school was removed even though that support is not part of the target capability. Or perhaps the student genuinely cannot yet explain the mechanism.

Those possibilities lead to very different teaching decisions.

The Opportunity Check asks whether the learner has been given a fair, relevant and sufficiently independent chance to perform the exact capability the tutor is trying to judge before the tutor turns a weak performance into a claim about missing capability.

This sounds like a technical assessment problem. In tutoring it is an everyday problem. The learner who “cannot solve algebra word problems” may have been shown only chapter-labelled exercises in which the method was announced in advance. The learner who “cannot write independently” may have been given a planning frame so detailed that the adult system made most of the important decisions. The learner who “cannot retrieve vocabulary” may never have been asked to retrieve it after a delay. The learner who “cannot explain aloud” may understand the idea but be blocked by a response channel that is not the skill the tutor actually cares about.

Quick answer

Before saying that a learner lacks a capability, check four things. First, was the target capability actually taught or reasonably available to the learner? Second, did the task genuinely require that capability rather than a neighbouring one? Third, did the learner have a usable response route, including legitimate access arrangements and non-answer-giving supports? Fourth, was the opportunity independent enough that success or failure can be attributed to the learner rather than to hidden adult help?

If any of those conditions fail, the next move is not automatically more practice. The tutor may need to redesign the task, restore an access support, remove an irrelevant barrier, create a changed-condition check, or teach the missing prerequisite before judging the learner again.

The standard is deliberately narrow: do not confuse a weak sample with a missing skill when the learner was not given a valid chance to show the skill in the first place.

1. What this volume owns

This volume owns the distinction between failure to demonstrate and failure of capability inside tutoring decisions. It sits between diagnosis and intervention. It is not a general accessibility manual, not a guide to formal examination accommodations, not a theory of assessment validity, and not a replacement for subject-specific teaching.

The wider mechanism owners remain elsewhere in the eduKate estate. How Learning Diagnosis Works owns the broader move from visible difficulty to plausible causes. How We Know Learning Has Really Held owns retention, changed conditions and independence as evidence of durable learning. How Learning Works owns the wider mechanism map. This article adds a particular tutor question: was the evidence opportunity itself good enough to support the conclusion we are about to make?

That question matters because tutoring is unusually responsive. A tutor can change a prompt, example, representation, timing rule or amount of help within seconds. That flexibility is valuable, but it also makes it easy to change the very conditions that determine what the learner had an opportunity to demonstrate.

2. Two different errors look the same from a distance

Imagine two learners who both leave a question blank.

Learner A has been taught the underlying concept, has practised it in several forms, understands the language of the question and has previously shown the prerequisite skills. On a fresh task with appropriate support conditions, A still cannot begin. That blank is meaningful evidence of a capability gap or an unresolved bottleneck.

Learner B has never been taught the relevant representation and has never seen the notation used in the task. B also leaves the question blank. The visible result is identical. The interpretation is not. Calling B “weak at the concept” would turn an opportunity problem into a learner label.

The opposite mistake also occurs. A tutor keeps changing the task until the learner eventually succeeds, then concludes that the learner had the capability all along. But if the successful version quietly removed the target operation, that conclusion is too generous. A learner who cannot select a method independently has not demonstrated method selection merely because the tutor named the method and the learner executed it accurately.

The Opportunity Check therefore protects both directions. It prevents underestimating learners because the task was unfair or irrelevant, and it prevents overestimating learners because the task was over-supported or no longer measured the intended capability.

3. A genuine opportunity has to match the target

A task can look academically respectable while asking the wrong question.

If the target is algebraic execution, giving a learner a complex word problem may add reading and representation demands that make the execution skill hard to observe. If the target is reading inference, requiring a long written response may add sentence-production demands that obscure whether the learner made the inference. If the target is scientific causal reasoning, asking for memorised terminology may sample vocabulary more strongly than mechanism understanding.

This does not mean every task should be simplified. Real performance is integrated. Eventually learners must combine reading, representation, retrieval, selection, execution and communication. The tutor’s responsibility is to know which layer a particular task is being used to judge.

During diagnosis, isolate when isolation clarifies the cause. During reintegration, deliberately add neighbouring demands and watch whether the repaired skill survives. During examination preparation, restore the real conditions that matter. A fair opportunity is therefore not necessarily an easy opportunity. It is an opportunity whose demands are appropriate to the claim the tutor intends to make.

4. Opportunity begins before the question is asked

Sometimes the learner has not had an opportunity because the necessary knowledge has not yet been built.

A Primary learner is asked to compare two fractions with unlike denominators. If equivalence has not been established, failure is not clean evidence of weak comparison. A Secondary learner is asked to write an argumentative paragraph using evidence, but the difference between evidence and example has never been taught. A Science learner is asked to explain evaporation at particle level before the relevant particle model has been introduced.

Effort cannot substitute for missing instruction. “Try harder” is not a diagnostic method. Nor should the tutor infer laziness from a learner who cannot perform a move that the learning route has not made available yet.

The tutor should ask: what had to be available before this task could become a fair sample? That may include subject knowledge, vocabulary, notation, a representation, a procedure, familiarity with the response format or an explicitly taught checking routine. If the prerequisite was never built, teach it. If it was built but is not retrievable, test that. If it is retrievable but does not transfer, change the task. Opportunity is a chain, not merely the moment when the worksheet reaches the desk.

5. The response channel can hide the capability

What a learner knows and how a learner is asked to express it are connected but not identical.

CAST’s UDL and Assessment resource makes a useful assessment-design point: educators should consider which actions are actually relevant to the construct being measured and which actions can be varied or supported to obtain a more accurate picture of learning. CAST’s current UDL Guidelines 3.0 likewise emphasise multiple means of action and expression because learners differ in how they navigate and communicate within learning environments.

For tutoring, the practical translation is careful rather than ideological. If handwriting is not the target, illegible or physically laborious handwriting should not automatically be treated as lack of conceptual understanding. If oral fluency is not the target, a learner may be allowed to sketch a causal chain before explaining it. If mental arithmetic is not the target, a calculator may sometimes remain available. But if the target is precisely handwriting, oral explanation, mental computation or unaided retrieval, changing the response channel can remove the skill the tutor intended to inspect.

The tutor must therefore preserve the construct while reducing irrelevant barriers. Flexibility is not the same as lowering the target.

6. Access support is not the same as answer support

This distinction is essential when judging independence.

An access support helps the learner encounter or communicate the task without supplying the intellectual answer. A screen reader, enlarged text, approved scribe, adapted keyboard or permitted rest break can change access while preserving the intended knowledge and reasoning. Ofqual’s 2026 guidance on Reasonable Adjustments explicitly states that adjustments should remove substantial disadvantage while the assessment continues to test the same knowledge, skills and understanding.

Answer support is different. A method hint can replace method selection. A sentence frame can replace part of paragraph planning. A model answer visible before the first attempt can replace retrieval and construction. A parent who translates the question into the exact mathematical relationship may remove the representation demand.

Do not remove legitimate access arrangements in the name of “true independence”. Independence means the learner carries the target intellectual work under appropriate conditions. It does not mean every human difference or disability support must disappear. The Opportunity Check asks what support changes access and what support changes the target operation.

7. Opportunity is different from exposure

A learner may have seen a concept many times without ever having to produce the target capability.

Reading ten worked examples is exposure. Choosing the method on an unlabeled problem is an opportunity to demonstrate selection. Highlighting vocabulary in a text is exposure. Using a word precisely in a new sentence is an opportunity to demonstrate productive knowledge. Watching a Science explanation is exposure. Reconstructing the mechanism from a changed observation is an opportunity to demonstrate causal understanding.

Tutors sometimes overestimate opportunity because the material was present. “We covered this” does not mean “the learner had to do this.” A student can complete a whole chapter while the chapter structure quietly announces the method for every problem. A learner can copy polished corrections without ever making the corrected decision on fresh work.

AERO’s Extend and Challenge guide is useful here. It emphasises providing opportunities for students to extend knowledge and demonstrate mastery through tasks that require retrieval and application in increasingly complex or integrated ways. The tutor’s job is to ask what the learner actually had to retrieve, choose, explain or execute—not merely what was visible on the page.

8. Opportunity has to be independent enough to read

A learner can receive a genuine task and still not receive a genuine opportunity if help arrives too quickly.

The tutor asks a question, waits two seconds, then gives a hint. The learner completes the problem. The adult feels productive. The evidence is weak. We do not know what would have happened if the learner had been allowed to search, represent, retrieve or make a first move.

Independence is not all-or-nothing. A diagnostic opportunity can declare the support boundary: “You may ask me to clarify a word, but I will not identify the method.” A writing opportunity can allow a dictionary while withholding a planning frame because vocabulary lookup is not the target but paragraph organisation is. A Science task can allow the original diagram while withholding the model explanation.

Write the condition down when it matters. Later, “correct” becomes interpretable because the tutor knows whether the learner used a cue, example, formula, device, reference or adult prompt. This is one reason the Configuration Snapshot records allowed support and the evidence rule.

9. The four-question Opportunity Check

A tutor does not need a formal assessment battery before every lesson. A practical Opportunity Check can be run in four questions.

Availability: Has the learner been taught or otherwise reasonably acquired the knowledge and prerequisite operations needed for this task?
Match: Does the task actually require the capability I intend to judge, or is another demand dominating the result?
Access: Can the learner perceive, navigate and respond to the task using legitimate supports that do not supply the target answer?
Independence: Is the opportunity sufficiently free of answer-giving help that success or failure says something meaningful about the learner?

If all four pass, the performance becomes stronger evidence. If one fails, repair the opportunity before repairing the learner.

This is not a validated diagnostic score. Do not turn the four questions into a proprietary number. Their value is logical: they force the tutor to inspect the conditions under which the evidence was produced.

10. Mathematics case: “She cannot solve word problems”

Consider a fictional learner, Alicia. Her tutor notices that she performs strongly on algebra exercises but struggles on algebra word problems. The first interpretation is representation weakness: she cannot turn language into equations.

The Opportunity Check asks whether the current tasks actually isolate that claim. The tutor notices that many word problems contain unfamiliar vocabulary and multi-clause sentence structures. Alicia often asks what one relational phrase means before she begins. On a revised task using the same mathematical structure with simpler language, she represents the relationship correctly. On another task, the original language remains but key quantities are visually separated; she succeeds again.

The tutor should not conclude that Alicia has no representation weakness. The evidence now supports a narrower hypothesis: language access is contributing materially to the apparent representation problem. The next diagnostic move can separate vocabulary, parsing and algebraic mapping more deliberately.

Now reverse the case. If the tutor says, “This is a simultaneous-equations question; let x be the first quantity and y the second,” Alicia may solve accurately. That success does not demonstrate independent representation. The tutor supplied the representation route. A fair opportunity must remove the hint once access language is controlled.

11. English case: “He cannot infer”

Imagine a learner, Beatrice, who gives weak inference answers. The tutor has been requiring two-sentence written responses under time pressure. The answers are short, vague and sometimes grammatically incomplete.

Before diagnosing inference, the tutor asks Beatrice to point to the relevant phrase and explain orally what it suggests. The reasoning is accurate. She can compare two interpretations and reject the weaker one. When she writes, however, sentence construction consumes so much attention that the inferential link collapses.

This does not mean the writing problem should be ignored. It means the tutor now has two jobs rather than one false job. Inference can be trained and checked using a response channel that exposes the reasoning. Written expression can be trained separately, then reintegrated so the learner can eventually demonstrate inference in the actual written format required by school.

The reverse error is equally important. If the tutor asks only multiple-choice inference questions and Beatrice chooses correctly, that may not show she can construct a defensible inference independently. Recognition supplied by options is a different opportunity from generating and justifying an interpretation. The tutor should change the opportunity before changing the learning claim.

12. Science case: “She knows the facts but cannot explain”

Ciara remembers the relevant Science vocabulary. On worksheets she can identify “evaporation”, “water vapour” and “heat energy”. Yet her explanations are disjointed. The tutor could conclude that her causal understanding is weak.

The Opportunity Check changes the task. First, Ciara is given an unfamiliar everyday observation and asked to draw the causal chain using arrows. She builds a coherent sequence. Next, she explains the diagram without using the textbook wording. The mechanism remains sound. Finally, she writes a concise answer and loses one important relationship because she is trying to reproduce formal phrasing.

The evidence suggests that causal understanding is stronger than the original written response showed. The live problem is now translation from internal model to examination-ready language.

Again, opportunity does not mean permanent simplification. If PSLE or school Science ultimately requires precise written explanation, the route must return to that condition. But the tutor can now teach the right bridge. A learner should not receive months of “more Science understanding” practice when the missing operation is how to express an already-usable model under the required response form.

13. Studying case: “She cannot plan independently”

Emily appears unable to plan a study week. Every Sunday the tutor asks her to make a plan. Every Sunday she waits. The tutor then suggests which subjects come first, where retrieval should sit and which tasks can be deferred.

After six weeks, the tutor says Emily still cannot plan independently.

But Emily has never actually had the opportunity. The adult asks for a plan and then takes over the first difficult decisions before Emily makes them.

The Opportunity Check changes the support boundary. Emily gets the fixed constraints—school deadlines, available evenings, sleep boundary, one maintenance requirement—and ten uninterrupted minutes to build a first plan. The tutor can ask clarifying questions afterward but does not select priorities for her.

Now the evidence becomes meaningful. If Emily creates a workable sequence, the old claim was partly an opportunity failure. If she overloads Monday, ignores two deadlines and cannot explain her priorities, the tutor has genuine planning evidence to work with. The next intervention can target prioritisation, estimation or contingency rather than a vague label of “poor independent study”.

14. In a three-student tutorial, opportunity must be individual

Small groups create a special opportunity problem. The tutor may ask a question to the group, receive one strong answer, then move on. Two students have heard the answer but never had to generate it. Later, their familiarity can look like learning.

Three-student tuition can solve this if the tutor deliberately protects individual opportunities. One learner can answer first while the others write privately. All three can choose a method before discussion. Each can produce a short explanation before peer comparison. A shared worked example can be followed by separate changed questions so that the tutor sees what each learner carries.

The point is not to turn every class into three parallel one-to-one lessons. Group learning has real advantages: comparison, peer explanation, alternate representations and visible differences in reasoning. The tutor simply needs to know when group exposure has replaced individual demonstration.

A useful class rule is: when the evidence will drive an individual route decision, collect at least one individual opportunity before relying on the group response. That keeps convenience from becoming diagnosis.

15. The tutor classifications see different opportunity failures

The Tutor Classification Model describes tutor functions rather than permanent rankings of people. The Opportunity Check changes each function in a different way.

A Homework Helper asks whether the learner truly had a chance to manage the task before an adult organised it. An Explainer asks whether the learner had a chance to reconstruct the idea after explanation. A Drill Builder asks whether practice actually required retrieval, selection or fluency rather than mere repetition. A Diagnostic Tutor inspects whether task conditions are creating a false symptom. A Route Designer ensures later stages contain genuine opportunities to use earlier learning. A Performance Coach checks whether realistic conditions are sampling the intended performance rather than irrelevant speed or unfamiliar format. A Learning Architect watches whether several adult systems have quietly removed the learner’s opportunity to own the process.

The function does not change the principle: before changing the learner model, inspect the opportunity that generated the evidence.

16. Repair, Alignment and Frontier need different opportunities

In Repair mode, the tutor often needs a narrow opportunity that isolates the weak link. If the learner is repairing fraction comparison, a carefully chosen task can remove unrelated reading demands and reveal whether the mathematical relationship is now available.

In Alignment mode, the opportunity must increasingly resemble current school demand. A learner may understand an idea under generous tutoring conditions but still need to demonstrate it in the response form, timing and mixed context expected in school. Alignment should therefore restore relevant real-world constraints deliberately rather than all at once.

In Frontier mode, opportunity means more than “harder questions”. The learner needs tasks that genuinely require transfer, synthesis, explanation or independent problem framing. Giving advanced-looking content while telling the learner exactly what method to use can create the appearance of extension without a corresponding increase in learner agency.

The mode changes what a fair opportunity should look like. Repair clarifies the component. Alignment restores the operating environment. Frontier widens the range of situations in which the learner must choose and create.

17. Parents: ask what the child was actually required to do

When a tutor says, “She can do it now,” a useful parent question is not immediately “What mark did she get?” Ask: what did she have to do without help?

When a tutor says, “He still cannot do it,” ask: what opportunity did he have to show it, and what other demands were inside that task?

These questions are not challenges to the tutor’s authority. They clarify the evidence. A strong tutor should be able to explain the target, support conditions and next discriminating check without pretending every performance is clean.

Parents should also notice home support. If a child completes homework only after the parent restates each question, chooses the first method and checks every step, the completed page is evidence of work under high support. That support may be entirely appropriate for learning. It simply should not be mistaken for unsupported capability.

Equally, do not remove a legitimate access arrangement because the family wants to know what the child can “really” do. The relevant question is whether the child performs the target intellectual operation, not whether every support disappears.

18. The research and policy foundation is narrower than the tutoring application

The Opportunity Check is an eduKate tutoring framework, not a validated diagnostic instrument. Its ingredients are supported by several established educational principles, but no cited source validates this exact four-question tutor protocol.

AERO’s Extend and Challenge guidance emphasises giving students opportunities to apply learning and demonstrate mastery through appropriately challenging tasks. CAST’s UDL guidance emphasises distinguishing the target from irrelevant response barriers and allowing multiple forms of action and expression where the medium is not itself the construct. AERA’s longstanding Position Statement on High-Stakes Testing warns against major decisions from a single score and notes that when a score may not adequately reflect true proficiency, acceptable alternative means of demonstrating attainment matter. Ofqual’s 2026 guidance on reasonable adjustments similarly protects access while preserving the knowledge, skills and understanding being assessed.

Those sources come from different countries and purposes. They should not be treated as proof of one Singapore tuition method. They support the underlying discipline: interpret performance in light of what the learner was actually given an opportunity to demonstrate.

19. A practical Opportunity Check card

Target: What exact capability am I trying to judge?
Prerequisites: What must already be known or available?
Opportunity: Has the learner actually had to perform this capability before?
Task match: What neighbouring demands might dominate the result?
Access: What legitimate support is needed to encounter or express the task?
Answer support: What help would replace part of the target operation?
Independence: How much uninterrupted learner work is needed before help?
Changed condition: What small variation would separate the leading explanations?
Claim: What can this attempt support, and what can it not support?
Next move: Teach, restore access, reduce irrelevant load, fade answer support, retest or integrate?

The card is deliberately descriptive. It should not become a score. The tutor’s judgement remains anchored to the learner’s actual work and the target being taught.

20. Common failure: “He had the worksheet, so he had the opportunity”

Possessing the worksheet is not the same as possessing the prerequisites, understanding the task, accessing the response channel or receiving enough independent time to act.

Another failure sounds more sophisticated: “I varied the task, so the evidence is valid.” Variation can introduce new barriers. A changed representation may test flexibility; it may also introduce notation the learner has never encountered. A timed task may test efficiency; it may also begin measuring speed more strongly than the target knowledge. A no-notes condition may test retrieval; it may also remove a reference that is legitimately part of the intended real-world task.

The Opportunity Check does not declare one universal testing condition. It requires the tutor to name the target and defend the conditions used to sample it.

21. Common failure: redesign the task until the learner succeeds

Fairness does not mean guaranteeing success.

If the target is independent method selection, eventually the learner must choose without a method cue. If the target is written argument, eventually the learner must produce the argument in writing. If the target is examination performance, eventually the relevant examination constraints have to return.

The tutor should remove irrelevant barriers, not the intellectual demand that defines the skill. This is why the opportunity check must always begin with the target. Without a target, every difficulty can look unfair and every success can look sufficient.

A good tutor can say, “I changed the response channel because handwriting is not what I am diagnosing today,” and later say, “Now we are restoring the written format because written production is part of the school task we need to align with.” Conditions are chosen for a reason, not because easier or harder is inherently better.

22. Common failure: remove all support to discover the “true learner”

There is no single support-free condition that reveals the learner’s essence.

Real education contains tools, language, diagrams, teachers, peers, references and accommodations. The relevant question is which supports belong to the target and which supports replace the target.

A learner who uses glasses is not less independent. A learner who uses an approved screen reader can still perform demanding reasoning. A learner who consults a formula sheet may still demonstrate modelling and execution if formula recall is not the capability under inspection. Conversely, a learner who receives the exact method from a tutor has not demonstrated independent method selection merely because the final calculation is correct.

Independence should be defined at the level of the intellectual job, not as the absence of every form of support.

23. What the Opportunity Check changes in a tutor’s language

Weak language says, “She cannot do inference.”

Stronger language says, “On two timed written responses, she did not produce a defensible inference. When the writing demand was reduced and the text remained available, her oral reasoning was accurate. I am separating inference from written expression before deciding which capability needs repair.”

Weak language says, “He only understands with help.”

Stronger language says, “He completes the procedure after a method cue. We do not yet have evidence that he selects the method independently. The next check removes the cue but keeps the vocabulary support.”

This language is longer because the evidence is more precise. Over time, the tutor becomes faster at making these distinctions. The goal is not to turn every lesson into a research report. The goal is to stop a poor opportunity from becoming a confident learner label.

24. The ethical standard

Children are unusually vulnerable to adult interpretation. A weak sample can quickly become a story: careless, weak in English, bad at problem solving, dependent, slow, not a Science child, not ready for Additional Mathematics.

Some of those stories begin with real difficulties. The ethical problem is not noticing difficulty. It is claiming more than the opportunity allowed us to see.

A tutor should not repair the learner until the tutor has checked whether the learning opportunity itself needs repair. Give the learner the right task, preserve legitimate access, remove answer-giving help where the target requires independence, allow enough space for the learner to act, and only then let the performance update the learner model.

That discipline protects the learner from two opposite harms: being underestimated because the opportunity was poor, and being overestimated because the adult quietly performed the hardest part.

Evidence and connected reading

Final compression

Before you say the learner cannot do it, ask whether the learner was actually given the chance to do it.

Check the prerequisite. Check the task. Check the response channel. Check the access support. Check the amount of answer help. Check the independence condition.

Then collect the performance.

Then make the claim.

The Opportunity Check protects a simple principle: a learner should be judged on evidence produced by a task that actually allowed the intended capability to appear.

That is the Opportunity Check.

That is Tutor Handbook Volume 0068.

Small Group Tutorials

The Tutor Handbook Vol No.0068 | The Opportunity Check — How a Tutor Distinguishes Missing Capability From a Missing Opportunity to Demonstrate It