eduKateSG · Study & Learning Methods Hub · Prerequisite Gap or Method-Selection Gap? · 2 October 2026
Two students can write the same wrong answer and need opposite repairs. One may be missing a prerequisite. The other may know every required method but fail to choose the right one when the problem arrives without a label. If the diagnosis is wrong, the teaching can become beautifully executed repair of the wrong mechanism.
This guide develops SG-AUD-20260929-03 beneath eduKateSG’s Study & Learning Methods Hub. It connects How Learning Dependencies Work | What Must Be Ready First with How Transfer of Learning Works | When Knowledge Survives a New Situation. The first asks whether upstream capability is ready. The second asks whether existing knowledge can be recognised and adapted under changed conditions.
The distinction is especially important after a mark drop. ‘Student cannot do topic X’ is not yet a diagnosis. The failure could come from missing factual knowledge, a weak representation, unavailable vocabulary, fragile retrieval, the wrong method choice, execution error, checking failure or transfer breakdown. The repair should follow the mechanism rather than the topic label.
This article remains a public diagnostic framework, not a personal assessment. Individual learner-state decisions belong in context, with learner voice and task evidence, and eduKateSG’s federation routes live learner-runtime work to Sengkang while subject specialists retain deep English and Mathematics validation.
The same wrong answer can come from different mechanisms
The same wrong answer can come from different mechanisms is mainly about refusing to diagnose from the surface error alone. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is two students both choosing the wrong algebra method for different reasons. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is assuming identical errors require identical teaching. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the diagnosis is based on follow-up evidence. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
A prerequisite gap means something necessary is not ready
A prerequisite gap means something necessary is not ready is mainly about locating missing knowledge or subskill upstream of the target task. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a student unable to simplify fractions while solving algebraic equations with rational coefficients. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is re-teaching equation strategy when fraction control is the true bottleneck. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the missing dependency fails even in a simpler isolated task. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
A method-selection gap means the methods exist but the chooser fails
A method-selection gap means the methods exist but the chooser fails is mainly about testing discrimination among known strategies. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a learner who can solve simultaneous equations and quadratic equations separately but chooses incorrectly in a mixed set. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is re-teaching both procedures from zero. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the learner executes each method when cued but misidentifies when to use it. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Retrieval failure is different again
Retrieval failure is different again is mainly about checking whether the learner knows but cannot bring the knowledge to mind. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a student recognising a formula in notes but unable to recall it during a fresh problem. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is calling the issue method selection before testing recall. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the diagnosis distinguishes unavailable knowledge from wrong choice. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Representation can be the hidden prerequisite
Representation can be the hidden prerequisite is mainly about checking whether the learner has formed the right problem model. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a word problem misread as additive because the relationship was never represented correctly. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is giving more calculation practice. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the learner can restate or diagram the structure before choosing a method. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Vocabulary can be a prerequisite to reasoning
Vocabulary can be a prerequisite to reasoning is mainly about recognising that language access can block subject performance. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a Science learner misunderstanding ‘increase by’ versus ‘increase to’. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is diagnosing weak Science concept knowledge immediately. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the key terms can be interpreted independently of the full problem. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Concept knowledge and procedure can separate
Concept knowledge and procedure can separate is mainly about checking whether steps are memorised without meaning. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a student performing percentage-change calculations on routine items but unable to explain the denominator. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is treating procedural fluency as conceptual readiness. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the learner can explain what each quantity represents. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Procedure and selection can also separate
Procedure and selection can also separate is mainly about recognising that accurate execution does not guarantee strategic choice. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a learner who completes a worked method correctly after the teacher names it. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is assuming method mastery includes selection. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that mixed tasks test whether the method is chosen independently. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
A minimal probe is better than a long diagnostic first
A minimal probe is better than a long diagnostic first is mainly about using the smallest task that distinguishes explanations. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is asking one fraction simplification item and one mixed-method item before assigning a full remedial set. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is giving a 50-question test before forming a hypothesis. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the probe changes which repair would be chosen. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Cue response is diagnostic evidence
Cue response is diagnostic evidence is mainly about observing what happens when a small hint is added. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a student solving correctly after being asked ‘what relationship do these two equations share?’. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is treating the prompted success as independent mastery. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the type of cue reveals what was missing. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Different prompts test different gaps
Different prompts test different gaps is mainly about using hints intentionally rather than randomly. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a vocabulary cue, representation cue and method cue producing different response patterns. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is helping until the learner succeeds and then forgetting which help mattered. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the prompt is tied to a diagnostic hypothesis. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
A dependency map prevents downstream repair
A dependency map prevents downstream repair is mainly about tracing what must be ready first. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is checking integer operations before solving linear equations with negative coefficients. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is drilling equations while sign errors persist. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that repair begins at the earliest unstable dependency. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Dependencies are not always a strict ladder
Dependencies are not always a strict ladder is mainly about allowing mutually reinforcing knowledge. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is learning ratio concepts while also practising representations that clarify them. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is refusing to teach anything downstream until every prerequisite is perfect. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the learner receives enough foundation and enough contextual use. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Method selection is a discrimination problem
Method selection is a discrimination problem is mainly about teaching the cues that separate similar strategies. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is comparing when to use area versus perimeter. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is giving more same-type blocked practice. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the learner can state the distinguishing cue. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Contrast can expose method boundaries
Contrast can expose method boundaries is mainly about placing similar problems side by side. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is comparing direct and inverse proportion examples. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is teaching each chapter weeks apart with no comparison. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the learner learns why one method does not apply to the other case. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Interleaving tests selection
Interleaving tests selection is mainly about removing chapter labels and mixing confusable methods. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a mixed Mathematics set after basic procedures are stable. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is keeping practice blocked indefinitely. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that strategy choice improves on unlabelled tasks. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Worked examples can diagnose missing structure
Worked examples can diagnose missing structure is mainly about asking learners to explain why each step was chosen. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a student understanding steps but not the initial equation setup. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is checking only whether the final answer is copied correctly. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the explanation reveals the first missing decision. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Error logs should classify mechanism
Error logs should classify mechanism is mainly about recording whether an error is knowledge, representation, selection, execution or checking. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is tagging repeated sign errors differently from wrong-method choices. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is keeping one undifferentiated list of mistakes. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that future practice is allocated to the right mechanism. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Speed is weak diagnostic evidence
Speed is weak diagnostic evidence is mainly about avoiding assumptions from slow performance alone. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a careful student selecting the right method slowly. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is calling slowness a prerequisite gap. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that accuracy, cue use and explanation are checked separately. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Confidence is also weak diagnostic evidence
Confidence is also weak diagnostic evidence is mainly about separating certainty from competence. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a confident learner repeatedly applying the wrong strategy. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is using confidence as evidence that the student knows the material. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the diagnosis relies on task evidence. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
A correct answer can hide a prerequisite gap
A correct answer can hide a prerequisite gap is mainly about checking whether success depended on a workaround. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a student using a calculator to avoid weak fraction understanding. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is assuming the correct result proves the foundation is stable. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the underlying component is tested directly when it matters. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
A wrong answer can hide good method selection
A wrong answer can hide good method selection is mainly about preserving what already works. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a learner choosing the correct simultaneous-equation method but making one arithmetic slip. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is re-teaching strategy selection because the answer is wrong. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the repair targets execution rather than replacing correct reasoning. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Repairing the wrong layer can create boredom
Repairing the wrong layer can create boredom is mainly about recognising the cost of unnecessary foundational review. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is making a capable student repeat basic procedures they already control. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is assuming remediation is always harmless. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that diagnostic evidence justifies how far back to rebuild. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Repairing too high can create frustration
Repairing too high can create frustration is mainly about recognising the cost of skipping prerequisites. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is assigning sophisticated mixed problems while basic representation is unstable. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is calling repeated failure a motivation problem. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the learner can succeed on appropriately simplified precursor tasks. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Hybrid gaps are common
Hybrid gaps are common is mainly about allowing more than one mechanism to be weak. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a student with shaky fraction skills and poor method discrimination. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is forcing every case into one diagnosis. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the repair sequence identifies which gap should be addressed first. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
The first weak link is not always the only weak link
The first weak link is not always the only weak link is mainly about retesting after repair. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is fixing integer operations and then discovering a separate equation-selection issue. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is assuming the original diagnosis explains every later difficulty. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the system is re-diagnosed after each major repair. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Transfer gaps can look like prerequisite gaps
Transfer gaps can look like prerequisite gaps is mainly about checking whether knowledge works only in familiar contexts. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a learner solving ratios in recipes but not map scales. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is re-teaching basic ratio facts immediately. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that a near-transfer task tests portability before foundation is rebuilt. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Transfer gaps can look like selection gaps
Transfer gaps can look like selection gaps is mainly about seeing when the learner has the right method but fails to recognise structural similarity. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a student not applying proportional reasoning because the context changed. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is assuming the student forgot the method. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that changed examples test recognition of deep structure. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Question wording can create apparent method-selection errors
Question wording can create apparent method-selection errors is mainly about checking language before strategy. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a learner misunderstanding ‘at least’ and choosing an incorrect probability setup. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is drilling probability methods only. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the language interpretation is tested separately. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Exam stress can disrupt selection
Exam stress can disrupt selection is mainly about separating knowledge from pressured access. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a student selecting accurately in untimed mixed work but poorly under time pressure. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is rebuilding the entire prerequisite chain. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that timed and untimed evidence are compared. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Tutor prompts can conceal selection weakness
Tutor prompts can conceal selection weakness is mainly about tracking how much method choice the tutor supplies. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is asking ‘should we use Pythagoras here?’ before the student decides. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is crediting successful execution as strategic independence. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the next task removes the method cue. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Homework structure can conceal selection weakness
Homework structure can conceal selection weakness is mainly about noticing when chapter labels tell the learner what to do. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a worksheet titled ‘Quadratic Formula Practice’. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is assuming high homework accuracy means flexible problem solving. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that cumulative unlabeled review is included. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Subject specialists should validate the repair
Subject specialists should validate the repair is mainly about routing deep disciplinary judgments to the correct expertise. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is using BTT for advanced Mathematics method distinctions or SETC for English language-performance mechanisms. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is turning a general learning-diagnostic page into a subject-specific authority. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that cross-domain diagnosis remains a framework with specialist handoffs. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Sengkang owns learner-runtime interpretation
Sengkang owns learner-runtime interpretation is mainly about preserving the federation boundary for individual learner decisions. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is routing a live student’s first-weak-link diagnosis into the Sengkang learner runtime. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is using a public article to prescribe an individual repair plan. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the page explains method while personal diagnosis stays contextual. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
A diagnostic hypothesis should predict the response to repair
A diagnostic hypothesis should predict the response to repair is mainly about making explanations testable. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is predicting that fraction repair will reduce algebra sign and coefficient errors if the dependency hypothesis is correct. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is choosing a repair without stating what should improve. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that later evidence can confirm or weaken the diagnosis. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Changed-condition retesting distinguishes repair types
Changed-condition retesting distinguishes repair types is mainly about checking whether improvement survives outside the training format. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a student practising fraction foundations then succeeding on fresh algebra items. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is testing only the exact remedial exercises. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the downstream target improves if the prerequisite repair was relevant. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Method-selection repair should improve choice before speed
Method-selection repair should improve choice before speed is mainly about expecting the correct sequence of change. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is a learner choosing the right strategy more often even while mixed-set completion remains slow. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is calling the intervention unsuccessful because time did not immediately fall. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that selection accuracy is checked before fluency. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Prerequisite repair should reduce recurring downstream noise
Prerequisite repair should reduce recurring downstream noise is mainly about looking for fewer low-level interruptions in advanced work. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is fewer fraction mistakes during algebra after a foundation rebuild. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is judging success only on isolated fraction drills. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the repaired dependency shows up in the target task. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Cross-subject evidence should preserve the subject’s own standards
Cross-subject evidence should preserve the subject’s own standards is mainly about using the framework without pretending every subject diagnoses errors the same way. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is English inference requiring evidence and scope while Mathematics method choice depends on structural cues. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is using one generic error taxonomy without disciplinary interpretation. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the general mechanism is translated into subject-valid evidence. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
The diagnostic sequence should be reversible
The diagnostic sequence should be reversible is mainly about avoiding labels that persist after the evidence changes. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is removing a ‘weak foundations’ hypothesis when fresh prerequisite checks become stable. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is letting an early diagnosis become the learner’s identity. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the explanation is updated when later evidence contradicts it. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
Learner explanation is itself a probe
Learner explanation is itself a probe is mainly about asking how the student decided what to do. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is hearing a student say they chose the method because the question contained one familiar word. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is inferring strategy from written work alone. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the learner’s reasoning helps distinguish knowledge from selection. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
The final goal is the smallest correct repair
The final goal is the smallest correct repair is mainly about avoiding both under-treatment and unnecessary rebuilding. The first diagnostic discipline is to resist the surface. A wrong answer is an outcome, not a mechanism. The teacher or tutor needs one or two well-chosen probes that make the competing explanations behave differently.
A practical example is teaching exactly the missing dependency or discrimination rule then returning to the real task. The example matters because the smallest useful diagnostic task is often simpler than the original failure. If the learner cannot perform the prerequisite in isolation, the dependency hypothesis becomes stronger. If the learner performs both candidate methods accurately when cued but fails in a mixed set, selection becomes the more plausible target.
The common failure is resetting an entire subject whenever marks fall. This wastes time and can damage calibration. Over-remediation tells capable learners they have forgotten foundations they still possess. Under-remediation keeps asking for advanced performance from a system missing an upstream component. Good diagnosis protects both challenge and support.
The acceptance test is that the learner returns to normal progression as soon as the bottleneck is stable. The repair should improve the downstream target under a fresh condition. If the expected change does not appear, revise the diagnosis. Learning problems are hypotheses about a system, not permanent labels attached to a student.
A six-probe diagnostic sequence
- Probe the suspected prerequisite in isolation.
- Ask the learner to explain or represent the target problem before solving it.
- Give the correct method name and see whether execution succeeds.
- Remove the method cue and mix the target with one plausible alternative.
- Change the surface context while keeping the underlying structure.
- Retest after the repair and check whether the downstream target improved.
Two contrasting Mathematics cases
Case A · Prerequisite gap
A student repeatedly fails linear equations containing fractional coefficients. On a simple fraction-equivalence task, the student also struggles. When the equation method is named, fraction errors still disrupt execution. The most economical hypothesis is an upstream fraction dependency. The repair begins there, then returns quickly to equations to verify downstream benefit.
Case B · Method-selection gap
A second student solves linear equations and simultaneous equations accurately when worksheets are separated by topic. In a mixed set, the student repeatedly uses one-equation methods on two-variable systems. The prerequisite procedures are present. The likely gap is discrimination: recognising structural cues that determine method choice. The repair is contrast and interleaving, not a return to arithmetic foundations.
The public diagnostic table
| Observed pattern | More consistent with prerequisite gap | More consistent with selection / transfer gap |
|---|---|---|
| Simpler component task | Fails | Often succeeds |
| Correct method supplied | Still fails because underlying component breaks | Often succeeds |
| Blocked same-type practice | May remain unstable | Often looks strong |
| Mixed practice | Fails through weak component and may look noisy | Selection errors become visible |
| Changed context | Underlying weakness persists | Transfer/recognition failure may appear |
| Best first repair | Rebuild earliest unstable dependency | Contrast, cues, interleaving and transfer practice |
Frequently asked questions
What is a prerequisite gap?
A prerequisite gap is a missing or unstable piece of knowledge or skill that the target task depends on. The weakness should usually appear even when the prerequisite is tested more simply.
What is a method-selection gap?
The learner can perform relevant methods when cued but has difficulty deciding which method fits an unlabelled problem. It is often a discrimination or transfer problem.
Can both gaps exist at the same time?
Yes. Hybrid cases are common. Repair the earliest or most constraining weakness, then re-diagnose rather than assuming one repair solves everything.
How can I tell without a long diagnostic test?
Use small probes: isolate the prerequisite, provide a method cue, mix two related problem types and change the context. The pattern of responses is often more informative than one long score.
Does a wrong method always mean method selection is weak?
No. A weak representation or vocabulary misunderstanding can make the learner misclassify the problem before method choice even begins.
Why does blocked practice hide selection problems?
Because the worksheet or chapter often tells the learner which method to use. The student practises execution while the selection decision has already been made by the page.
What should I do if the learner succeeds only with prompts?
Record the prompt type. A representation cue, vocabulary cue or method cue provides diagnostic information. Then fade the prompt and test independence.
Can a high-performing student have a prerequisite gap?
Yes. Strong learners can build workarounds that hide weak foundations until task complexity rises. Test the suspected component directly rather than inferring from overall grades.
Can a low-performing student have good prerequisites?
Yes. Some learners know the components but struggle to recognise when and how to combine them. Their repair may need contrast, transfer and strategy selection rather than more basic drills.
When should specialist input be used?
When the distinction depends on deep subject knowledge, specialist English, Mathematics or Science expertise should validate the content-specific diagnosis. The framework does not replace disciplinary judgment.
The quiet conclusion: repair begins by asking what failed first
The visible mistake is often downstream. A learner writes the wrong equation, chooses the wrong evidence, uses the wrong tense or applies the wrong formula. The temptation is to repair the final step. Sometimes that is correct. Sometimes the final step is only where an earlier weakness became visible.
Good diagnosis works backwards just far enough. Test the dependency. Test the representation. Supply the method cue. Remove it. Change the context. Then choose the smallest repair that explains the evidence and return to the real task quickly. The goal is not to find a label for the learner. It is to find the first change that makes the system work again.