Full chapter index

Count chairs as markers and the spaces between them as intervals. Start with the concrete case: Six chairs stand in one straight row: ●—●—●—●—●—●. A child points to each chair and says ‘one gap’ six times. This is useful because it shows the child’s current model before an adult supplies a rule. Ask the child to describe what is being counted, read, measured or compared, and keep the original task visible while they speak.

The central idea is this: The first chair starts the row and the last chair ends it. A gap is created only when a new neighbouring chair is added after the first. The explanation should connect the surface feature to the underlying relationship. It should also say which conditions matter. A statement that gives only a result may earn a correct tick once, but it will not travel reliably to a changed example.

Use this teaching move: Mark the chairs 1 to 6, then mark the spaces A to E. Read the alternating sequence aloud. Let the child carry out the move and narrate the reason for each step. If a prompt is needed, give the smallest prompt that restarts thinking. Avoid completing the decisive step for them, because that makes fluent adult performance look like student understanding.

Now diagnose alternatives. If the child succeeds with the representation but fails without it, the concept is developing and memory load is still high. If the child cannot explain the representation, return to meaning. If the explanation is sound but the final answer is wrong, inspect execution, labels and checking rather than reteaching the whole topic.

A useful check is: The child should state, ‘Six endpoint markers make five internal intervals in this straight row.’ Ask for one near example and one changed example. The near example confirms the immediate repair; the changed example tests transfer. Keep a short note of the clue the child used, because progress is better shown by increasingly independent decisions than by repeating the same familiar item.

For parent support, finish with a calm question: ‘What would you look for first next time?’ The answer should name evidence, not a slogan. Praise the act of checking and the quality of the explanation. Then stop while the child is still accurate, so the lesson ends with a reliable method rather than fatigue-driven guessing.

This chapter focuses on a small decision with a large effect: Derive the relationship instead of presenting a chant. Consider the case carefully: Begin with one chair: there are no between-chair gaps. Add chairs one at a time and record pairs (1,0), (2,1), (3,2), up to (6,5). Before correcting anything, ask the child to predict what the task expects and to point to the clue supporting that prediction.

What is happening underneath is more precise than it first appears. Each added chair extends the row and creates exactly one new adjacency. Therefore gaps = chairs − 1 for this arrangement. That relationship is the part worth remembering. Names, pictures and numbers may change, but the same structure can be recovered when the learner asks what each element represents and how the elements are connected.

Try the following repair: Ask the child to predict ten chairs, then return to the growth pattern as proof. Work slowly enough for the child to see the decision point. Then remove one support—a label, a diagram, a prompt or a worked first line—and see whether the method survives. This gradual release distinguishes genuine learning from temporary imitation.

Compare two error routes. One learner may misunderstand the concept; another may understand but rush the notation or unit. Give each learner a different next task. The first needs a simpler contrast that makes the meaning visible. The second needs a checking routine that targets the exact execution risk.

The success criterion is not merely the printed answer. If the arrangement changes, the proof must be reconsidered; the formula belongs to conditions, not to the word ‘chair’. Invite the child to challenge the answer with a counterexample, substitution, diagram, rereading or repeat measurement as appropriate. When a method withstands that check, confidence becomes evidence-based rather than dependent on an adult’s reassurance.

At home, use one fresh example and one short reflection. Ask what changed, what stayed the same and which clue mattered most. That conversation keeps practice purposeful. It also helps a parent decide whether the difficulty is isolated, recurring across contexts, or already fading with a modest amount of guided practice.

The practical goal here is to make the hidden choice visible. Separate counting mistakes from representation and operation mistakes. Use this situation: Show seven dots, a fence with seven posts, and a number line from 0 to 6. Ask for objects, intervals and total equal lengths. Have the child mark the information supplied, the information requested and any boundary or condition before touching the calculation or final response.

A sound explanation follows the mechanism. A child may count inaccurately, ignore an endpoint, confuse labels with intervals, or know the structure but select the wrong operation. Encourage the learner to use because, therefore and only when accurately. Those connecting words expose missing reasoning: if the child cannot complete the causal chain, the final answer may be remembered rather than understood.

The next move should be concrete and repeatable. Have the child trace spaces with a finger, draw bars between dots and explain why the first dot does not create a preceding internal gap. After the first attempt, alter one feature while keeping the core relationship. That variation prevents the child from associating the method with one picture or one sentence pattern.

Watch how help changes the response. Immediate success after a leading question may show recognition, not independent recall. Wait briefly, ask for a plan, and offer a neutral prompt before a directional hint. Record which level of help was needed so tomorrow’s practice begins at the right point.

Use this criterion to decide whether to move on: Record which representation repairs the error fastest; that is the starting route for practice. The learner should also identify one tempting wrong route and explain why it fails. Contrasting a correct and incorrect interpretation develops error detection, which is often more durable than simply completing another similar item.

For a parent, the manageable next step is a three-item set: one familiar example, one changed example and one explanation question. If all three are secure, pause. If only the explanation is weak, discuss language. If the structure itself collapses, return to a simpler representation before adding speed.

Begin by narrowing the question. Use endpoints before applying spacing formulas. The case is: A 20 m straight fence has posts every 5 m, including a post at both ends. Ask the child to restate it without pronouns such as ‘it’ or vague words such as ‘thing’. Precise restatement frequently reveals whether the difficulty lies in reading, concept, representation or execution.

The dependable relationship is: The distance contains 20 ÷ 5 = 4 equal intervals. Four intervals need five boundary posts: at 0, 5, 10, 15 and 20 m. Link every technical term to an observable feature or a valid algebraic step. A child should be able to move in both directions—from the situation to the concept and from the concept back to a prediction about the situation.

A productive repair is: Draw the number line and write interval count first, then add one for the two-ended straight arrangement. Model it once only if necessary, then reset with changed values, wording or layout. Ask the child to decide before calculating. Prediction forces the learner to use a model, while calculation alone can sometimes conceal a guessed operation.

Checking should be targeted. A generic instruction to ‘check your work’ is too broad for many learners. Name the likely risk: boundary, sign, unit, scope, reference point, variable, viewpoint or conclusion. Later, fade the named risk so the learner selects the check independently.

The evidence of understanding is: Check by listing positions. A formula answer without endpoint positions can hide an off-by-one error. Ask why the check is capable of catching this particular mistake. That extra question prevents ritual checking, where a child repeats the same method and reproduces the same error with greater confidence.

Keep the emotional message light. An unusual answer or a corrected method is information, not a verdict on ability. The parent’s role is to preserve the question, invite an explanation and notice whether the child can use feedback on a new item. That is a more useful signal than speed on the first attempt.

Here the child needs a route, not a warning. Notice when the question counts spaces beside the row as well as spaces within it. Use the case: Six chairs are placed between two walls, and the question asks for every space from the left wall to the right wall. Ask for a drawing, paraphrase, table, labelled line or equipment check that matches the subject. The representation should clarify the task, not add decoration.

The reasoning can be stated clearly. Now the walls are boundaries. Depending on whether chairs touch the walls, there may be seven spaces: wall-to-chair, five chair-to-chair spaces, and chair-to-wall. Break that statement into a condition, a relationship and a consequence. Children often remember the consequence but omit the condition, which is why a rule works on one worksheet and fails on the next.

Guide the learner with this action: Draw every boundary explicitly and label which spaces the wording includes. When the child completes it, ask which part supplied new information. If they cannot answer, they may be following steps without understanding their purpose. Revisit the comparison that made the relationship visible.

Next, change the context while preserving the structure. Use different nouns, numbers, diagram orientation or measurement display. Ask the child to say why the new problem belongs to the same family. This classification step is central to transfer and reduces dependence on keywords.

Accept the work as secure when this is true: The word ‘gap’ is not enough; the start and end conditions decide the count. Then invite a concise final statement in the language of the question. Clear conclusions prevent a page of correct working from ending with an unlabeled number or an unexplained claim.

A parent can support without becoming the answer key. Ask for the plan, allow wait time and request one check. If the child is stuck, reveal a representation rather than the result. Finish by asking the child to create a small example, because creation shows which features they believe are essential.

This part builds independence around a predictable risk. Replace the straight-line rule when the last object neighbours the first. Examine the case: Six chairs are evenly placed around a round table. How many spaces are there between neighbouring chairs? Let the learner commit to an interpretation before seeing feedback. That commitment makes the later comparison informative and reduces the habit of changing answers merely because an adult sounds certain.

The subject knowledge behind the case is: The circle closes. Every chair has a next chair, including the sixth back to the first, so there are six intervals. Keep the wording exact enough to be testable. Avoid absolute language when conditions vary, and avoid vague language when the evidence supports a definite relationship.

Use a short cycle: Draw a loop and count each connecting arc once. Ask the learner to perform the cycle again with one detail changed. A reliable method should adapt; a memorised script often breaks when the appearance changes.

If an error appears, trace it to the earliest unsupported decision. Later arithmetic or wording may be perfectly consistent with that first wrong choice. Repairing the earliest decision is faster and kinder than marking every downstream line as a separate failure.

The final standard is: A child who automatically subtracts one has memorised a surface rule; the repair is to inspect whether the arrangement has endpoints. Ask the child to compare the original and changed cases, naming both the invariant structure and the feature that changes the answer. That comparison turns correction into generalisation.

For ongoing practice, space examples across several days and mix them with already-secure material. Retrieval after a delay is a better test than immediate repetition. A parent should look for shorter explanations that remain accurate, quicker selection of the right check and less need for rescuing prompts.

Move between number of intervals, interval length and total distance. Start with the concrete case: Nine lamp posts stand in a straight line, 4 m apart, with the first and ninth as endpoints. This is useful because it shows the child’s current model before an adult supplies a rule. Ask the child to describe what is being counted, read, measured or compared, and keep the original task visible while they speak.

The central idea is this: Nine posts create eight intervals, so the total distance is 8 × 4 = 32 m, not 9 × 4 m. The explanation should connect the surface feature to the underlying relationship. It should also say which conditions matter. A statement that gives only a result may earn a correct tick once, but it will not travel reliably to a changed example.

Use this teaching move: Write the units beside each quantity: posts, intervals and metres per interval. Let the child carry out the move and narrate the reason for each step. If a prompt is needed, give the smallest prompt that restarts thinking. Avoid completing the decisive step for them, because that makes fluent adult performance look like student understanding.

Now diagnose alternatives. If the child succeeds with the representation but fails without it, the concept is developing and memory load is still high. If the child cannot explain the representation, return to meaning. If the explanation is sound but the final answer is wrong, inspect execution, labels and checking rather than reteaching the whole topic.

A useful check is: The final check asks whether multiplying the number of posts would add a nonexistent length beyond the last post. Ask for one near example and one changed example. The near example confirms the immediate repair; the changed example tests transfer. Keep a short note of the clue the child used, because progress is better shown by increasingly independent decisions than by repeating the same familiar item.

For parent support, finish with a calm question: ‘What would you look for first next time?’ The answer should name evidence, not a slogan. Praise the act of checking and the quality of the explanation. Then stop while the child is still accurate, so the lesson ends with a reliable method rather than fatigue-driven guessing.

This chapter focuses on a small decision with a large effect: Compare an interval structure with an action structure. Consider the case carefully: One ribbon is cut into six pieces with separate straight cuts and no folding or stacking. Before correcting anything, ask the child to predict what the task expects and to point to the clue supporting that prediction.

What is happening underneath is more precise than it first appears. The uncut ribbon begins as one piece. Each cut increases the number of pieces by one, so six pieces require five cuts. That relationship is the part worth remembering. Names, pictures and numbers may change, but the same structure can be recovered when the learner asks what each element represents and how the elements are connected.

Try the following repair: Act it out with a strip of paper or draw the state after each cut. Work slowly enough for the child to see the decision point. Then remove one support—a label, a diagram, a prompt or a worked first line—and see whether the method survives. This gradual release distinguishes genuine learning from temporary imitation.

Compare two error routes. One learner may misunderstand the concept; another may understand but rush the notation or unit. Give each learner a different next task. The first needs a simpler contrast that makes the meaning visible. The second needs a checking routine that targets the exact execution risk.

The success criterion is not merely the printed answer. State the condition: one ribbon, sequential full cuts, no stacking. Different tools or arrangements can change the action count. Invite the child to challenge the answer with a counterexample, substitution, diagram, rereading or repeat measurement as appropriate. When a method withstands that check, confidence becomes evidence-based rather than dependent on an adult’s reassurance.

At home, use one fresh example and one short reflection. Ask what changed, what stayed the same and which clue mattered most. That conversation keeps practice purposeful. It also helps a parent decide whether the difficulty is isolated, recurring across contexts, or already fading with a modest amount of guided practice.

The practical goal here is to make the hidden choice visible. Decide whether the problem asks for stops visited or journeys between stops. Use this situation: A bus travels from Stop 1 to Stop 6 along one route segment. Have the child mark the information supplied, the information requested and any boundary or condition before touching the calculation or final response.

A sound explanation follows the mechanism. It visits six stops but makes five between-stop journeys. If a fare or travel time applies to each segment, use five intervals. Encourage the learner to use because, therefore and only when accurately. Those connecting words expose missing reasoning: if the child cannot complete the causal chain, the final answer may be remembered rather than understood.

The next move should be concrete and repeatable. List the journeys 1→2, 2→3, 3→4, 4→5 and 5→6. After the first attempt, alter one feature while keeping the core relationship. That variation prevents the child from associating the method with one picture or one sentence pattern.

Watch how help changes the response. Immediate success after a leading question may show recognition, not independent recall. Wait briefly, ask for a plan, and offer a neutral prompt before a directional hint. Record which level of help was needed so tomorrow’s practice begins at the right point.

Use this criterion to decide whether to move on: The check is semantic: a stop is a location; a segment is movement between two locations. The learner should also identify one tempting wrong route and explain why it fails. Contrasting a correct and incorrect interpretation develops error detection, which is often more durable than simply completing another similar item.

For a parent, the manageable next step is a three-item set: one familiar example, one changed example and one explanation question. If all three are secure, pause. If only the explanation is weak, discuss language. If the structure itself collapses, return to a simpler representation before adding speed.

Begin by narrowing the question. Apply the same endpoint logic to dates and scheduled events carefully. The case is: A plant is measured every two days on Days 1, 3, 5, 7 and 9. Ask the child to restate it without pronouns such as ‘it’ or vague words such as ‘thing’. Precise restatement frequently reveals whether the difficulty lies in reading, concept, representation or execution.

The dependable relationship is: There are five measurements but four two-day intervals between the first and last measurement, giving an elapsed time of 8 days. Link every technical term to an observable feature or a valid algebraic step. A child should be able to move in both directions—from the situation to the concept and from the concept back to a prediction about the situation.

A productive repair is: Separate event count from elapsed intervals and write the actual dates. Model it once only if necessary, then reset with changed values, wording or layout. Ask the child to decide before calculating. Prediction forces the learner to use a model, while calculation alone can sometimes conceal a guessed operation.

Checking should be targeted. A generic instruction to ‘check your work’ is too broad for many learners. Name the likely risk: boundary, sign, unit, scope, reference point, variable, viewpoint or conclusion. Later, fade the named risk so the learner selects the check independently.

The evidence of understanding is: This prevents children from multiplying five measurements by two days and adding time beyond the last observation. Ask why the check is capable of catching this particular mistake. That extra question prevents ritual checking, where a child repeats the same method and reproduces the same error with greater confidence.

Keep the emotional message light. An unusual answer or a corrected method is information, not a verdict on ability. The parent’s role is to preserve the question, invite an explanation and notice whether the child can use feedback on a new item. That is a more useful signal than speed on the first attempt.

Here the child needs a route, not a warning. Keep counts and measures in separate columns. Use the case: Problems may provide 12 trees, 3 m spacing and ask for total length; or provide 33 m and 3 m spacing and ask for trees. Ask for a drawing, paraphrase, table, labelled line or equipment check that matches the subject. The representation should clarify the task, not add decoration.

The reasoning can be stated clearly. The relationship links objects, intervals, size per interval and total measure, but these are not interchangeable units. Break that statement into a condition, a relationship and a consequence. Children often remember the consequence but omit the condition, which is why a rule works on one worksheet and fails on the next.

Guide the learner with this action: Use four headings: markers, intervals, measure per interval, total measure. Fill only known values before calculating. When the child completes it, ask which part supplied new information. If they cannot answer, they may be following steps without understanding their purpose. Revisit the comparison that made the relationship visible.

Next, change the context while preserving the structure. Use different nouns, numbers, diagram orientation or measurement display. Ask the child to say why the new problem belongs to the same family. This classification step is central to transfer and reduces dependence on keywords.

Accept the work as secure when this is true: A sensible answer must carry the requested unit and satisfy the arrangement conditions. Then invite a concise final statement in the language of the question. Clear conclusions prevent a page of correct working from ending with an unlabeled number or an unexplained claim.

A parent can support without becoming the answer key. Ask for the plan, allow wait time and request one check. If the child is stuck, reveal a representation rather than the result. Finish by asking the child to create a small example, because creation shows which features they believe are essential.

This part builds independence around a predictable risk. Build a reliable pre-calculation routine. Examine the case: Before solving, ask whether there is a marker at the start, a marker at the end, both, neither, or a closed loop. Let the learner commit to an interpretation before seeing feedback. That commitment makes the later comparison informative and reduces the habit of changing answers merely because an adult sounds certain.

The subject knowledge behind the case is: These five possibilities explain many apparent exceptions to the subtract-one rule. Keep the wording exact enough to be testable. Avoid absolute language when conditions vary, and avoid vague language when the evidence supports a definite relationship.

Use a short cycle: Sketch the boundary marks first, then decide how many intervals touch them. Ask the learner to perform the cycle again with one detail changed. A reliable method should adapt; a memorised script often breaks when the appearance changes.

If an error appears, trace it to the earliest unsupported decision. Later arithmetic or wording may be perfectly consistent with that first wrong choice. Repairing the earliest decision is faster and kinder than marking every downstream line as a separate failure.

The final standard is: The checklist turns a memory trick into structural reasoning and works even when the story uses trees, stations or decorative lights. Ask the child to compare the original and changed cases, naming both the invariant structure and the feature that changes the answer. That comparison turns correction into generalisation.

For ongoing practice, space examples across several days and mix them with already-secure material. Retrieval after a delay is a better test than immediate repetition. A parent should look for shorter explanations that remain accurate, quicker selection of the right check and less need for rescuing prompts.

Show why distance divided by spacing gives intervals, not always objects. Start with the concrete case: A 24 m path has markers every 6 m, including both ends. A child writes 24 ÷ 6 = 4 markers. This is useful because it shows the child’s current model before an adult supplies a rule. Ask the child to describe what is being counted, read, measured or compared, and keep the original task visible while they speak.

The central idea is this: The quotient counts four equal lengths. Boundaries of four consecutive lengths are at five positions. The explanation should connect the surface feature to the underlying relationship. It should also say which conditions matter. A statement that gives only a result may earn a correct tick once, but it will not travel reliably to a changed example.

Use this teaching move: Add positions 0, 6, 12, 18 and 24 m to verify. Let the child carry out the move and narrate the reason for each step. If a prompt is needed, give the smallest prompt that restarts thinking. Avoid completing the decisive step for them, because that makes fluent adult performance look like student understanding.

Now diagnose alternatives. If the child succeeds with the representation but fails without it, the concept is developing and memory load is still high. If the child cannot explain the representation, return to meaning. If the explanation is sound but the final answer is wrong, inspect execution, labels and checking rather than reteaching the whole topic.

A useful check is: The child should label the quotient ‘intervals’ before converting it to the requested object count. Ask for one near example and one changed example. The near example confirms the immediate repair; the changed example tests transfer. Keep a short note of the clue the child used, because progress is better shown by increasingly independent decisions than by repeating the same familiar item.

For parent support, finish with a calm question: ‘What would you look for first next time?’ The answer should name evidence, not a slogan. Praise the act of checking and the quality of the explanation. Then stop while the child is still accurate, so the lesson ends with a reliable method rather than fatigue-driven guessing.

This chapter focuses on a small decision with a large effect: Use counterexamples to protect flexible thinking. Consider the case carefully: A necklace has eight equally spaced beads in a loop, and a square garden has one post at each corner. Before correcting anything, ask the child to predict what the task expects and to point to the clue supporting that prediction.

What is happening underneath is more precise than it first appears. Closed paths return to the starting marker, so the number of intervals can equal the number of markers. Corners and sides also pair one-to-one in a simple polygon. That relationship is the part worth remembering. Names, pictures and numbers may change, but the same structure can be recovered when the learner asks what each element represents and how the elements are connected.

Try the following repair: Ask what neighbours the final object and whether a final-to-first interval exists. Work slowly enough for the child to see the decision point. Then remove one support—a label, a diagram, a prompt or a worked first line—and see whether the method survives. This gradual release distinguishes genuine learning from temporary imitation.

Compare two error routes. One learner may misunderstand the concept; another may understand but rush the notation or unit. Give each learner a different next task. The first needs a simpler contrast that makes the meaning visible. The second needs a checking routine that targets the exact execution risk.

The success criterion is not merely the printed answer. A counterexample is useful because it forces the child to state the condition behind the straight-line rule. Invite the child to challenge the answer with a counterexample, substitution, diagram, rereading or repeat measurement as appropriate. When a method withstands that check, confidence becomes evidence-based rather than dependent on an adult’s reassurance.

At home, use one fresh example and one short reflection. Ask what changed, what stayed the same and which clue mattered most. That conversation keeps practice purposeful. It also helps a parent decide whether the difficulty is isolated, recurring across contexts, or already fading with a modest amount of guided practice.

The practical goal here is to make the hidden choice visible. Vary one structural feature at a time. Use this situation: Start with straight rows with both endpoints, then remove one endpoint, add outside gaps, close the row into a circle and finally mix unknown distance or spacing. Have the child mark the information supplied, the information requested and any boundary or condition before touching the calculation or final response.

A sound explanation follows the mechanism. Random mixed worksheets can reward guessing if the child has not named the arrangement. Encourage the learner to use because, therefore and only when accurately. Those connecting words expose missing reasoning: if the child cannot complete the causal chain, the final answer may be remembered rather than understood.

The next move should be concrete and repeatable. Require a tiny sketch, labels and a sentence identifying what the first calculation counts. After the first attempt, alter one feature while keeping the core relationship. That variation prevents the child from associating the method with one picture or one sentence pattern.

Watch how help changes the response. Immediate success after a leading question may show recognition, not independent recall. Wait briefly, ask for a plan, and offer a neutral prompt before a directional hint. Record which level of help was needed so tomorrow’s practice begins at the right point.

Use this criterion to decide whether to move on: Fade the sketch only after the child can reconstruct it mentally and explain exceptions. The learner should also identify one tempting wrong route and explain why it fails. Contrasting a correct and incorrect interpretation develops error detection, which is often more durable than simply completing another similar item.

For a parent, the manageable next step is a three-item set: one familiar example, one changed example and one explanation question. If all three are secure, pause. If only the explanation is weak, discuss language. If the structure itself collapses, return to a simpler representation before adding speed.

Begin by narrowing the question. Look for diagnosis and transfer rather than a slogan. The case is: A child may answer chair questions correctly yet fail lamp-post questions because the story changed. Ask the child to restate it without pronouns such as ‘it’ or vague words such as ‘thing’. Precise restatement frequently reveals whether the difficulty lies in reading, concept, representation or execution.

The dependable relationship is: Effective tuition should compare isomorphic problems, change endpoints deliberately and ask the child to generate an example where objects equal intervals. Link every technical term to an observable feature or a valid algebraic step. A child should be able to move in both directions—from the situation to the concept and from the concept back to a prediction about the situation.

A productive repair is: Use unfamiliar contexts after guided examples and require a check with positions or a diagram. Model it once only if necessary, then reset with changed values, wording or layout. Ask the child to decide before calculating. Prediction forces the learner to use a model, while calculation alone can sometimes conceal a guessed operation.

Checking should be targeted. A generic instruction to ‘check your work’ is too broad for many learners. Name the likely risk: boundary, sign, unit, scope, reference point, variable, viewpoint or conclusion. Later, fade the named risk so the learner selects the check independently.

The evidence of understanding is: Mastery is visible when the child selects the relationship from structure, not from a keyword. Ask why the check is capable of catching this particular mistake. That extra question prevents ritual checking, where a child repeats the same method and reproduces the same error with greater confidence.

Keep the emotional message light. An unusual answer or a corrected method is information, not a verdict on ability. The parent’s role is to preserve the question, invite an explanation and notice whether the child can use feedback on a new item. That is a more useful signal than speed on the first attempt.

Here the child needs a route, not a warning. Choose support according to the pattern of errors. Use the case: Parents ask whether the child is careless, whether a formula should be memorised, and whether bar models are required. Ask for a drawing, paraphrase, table, labelled line or equipment check that matches the subject. The representation should clarify the task, not add decoration.

The reasoning can be stated clearly. One off-by-one mistake may need only a concrete row. Repeated errors across time, measurement and patterns suggest a broader representation gap. Break that statement into a condition, a relationship and a consequence. Children often remember the consequence but omit the condition, which is why a rule works on one worksheet and fails on the next.

Guide the learner with this action: Keep formulas as summaries after derivation; use any clear diagram that shows markers, intervals and endpoints. When the child completes it, ask which part supplied new information. If they cannot answer, they may be following steps without understanding their purpose. Revisit the comparison that made the relationship visible.

Next, change the context while preserving the structure. Use different nouns, numbers, diagram orientation or measurement display. Ask the child to say why the new problem belongs to the same family. This classification step is central to transfer and reduces dependence on keywords.

Accept the work as secure when this is true: Consider sustained support when the child cannot transfer after varied guided examples, not merely after one unusual puzzle. Then invite a concise final statement in the language of the question. Clear conclusions prevent a page of correct working from ending with an unlabeled number or an unexplained claim.

A parent can support without becoming the answer key. Ask for the plan, allow wait time and request one check. If the child is stuck, reveal a representation rather than the result. Finish by asking the child to create a small example, because creation shows which features they believe are essential.

This part builds independence around a predictable risk. Finish with independent reasoning across contexts. Examine the case: Day 1 uses coins; Day 2 uses fence posts; Day 3 uses distance and spacing; Day 4 uses cuts; Day 5 uses a circle; Day 6 mixes cases; Day 7 asks the child to teach the rule and its exceptions. Let the learner commit to an interpretation before seeing feedback. That commitment makes the later comparison informative and reduces the habit of changing answers merely because an adult sounds certain.

The subject knowledge behind the case is: Keep sessions short and ask for labels before arithmetic. Keep the wording exact enough to be testable. Avoid absolute language when conditions vary, and avoid vague language when the evidence supports a definite relationship.

Use a short cycle: Reduce prompts from a full endpoint checklist to one question: ‘What exactly are you counting?’ Ask the learner to perform the cycle again with one detail changed. A reliable method should adapt; a memorised script often breaks when the appearance changes.

If an error appears, trace it to the earliest unsupported decision. Later arithmetic or wording may be perfectly consistent with that first wrong choice. Repairing the earliest decision is faster and kinder than marking every downstream line as a separate failure.

The final standard is: The goal is a child who can model a new interval problem, justify the relationship and catch an extra or missing count independently. Ask the child to compare the original and changed cases, naming both the invariant structure and the feature that changes the answer. That comparison turns correction into generalisation.

For ongoing practice, space examples across several days and mix them with already-secure material. Retrieval after a delay is a better test than immediate repetition. A parent should look for shorter explanations that remain accurate, quicker selection of the right check and less need for rescuing prompts.