eduKateSG · Cambridge Primary English · Cambridge Primary Stage 3
Find the next useful language move
You are here to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it.
Open the grouped contents · Go to practice and repair · Use the Vocabulary Learning Hub
The Core Aim of Cambridge Primary Stage 3 Vocabulary | Comparing Two Information Cards Without Mixing Their Facts is for a learner working within Cambridge Primary Stage 3 English who can read short factual texts but sometimes merges details while comparing them. The reader is here because a familiar classroom task can become unexpectedly difficult when the words for order, evidence, comparison or decision are not yet available on demand.
The useful answer is immediate: hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it. That aim links vocabulary to comprehension, retrieval, speaking, writing and age-appropriate independence. It does not require a longer word list; it requires a small set of words whose meanings and jobs are understood deeply enough to travel.
Cambridge Primary Stage 3 is used as an official curriculum-stage label. This article does not convert it into a Singapore, British, American or other school level, and the cards are original teaching texts rather than Cambridge assessment materials. Cambridge describes Cambridge Primary as a flexible programme typically serving learners aged 5 to 11, and its current Primary English materials span Stages 1 to 6. The stage label in this article is therefore kept intact rather than translated into another system.
Five reading routes
Chapters 1–6
Define the learner job and locate the earliest unstable link.
Chapters 7–12
Strengthen distinctions, reference, grammar and useful word partnerships.
Chapters 13–18
Move from guided language into the complete classroom product.
Chapters 19–25
Use original cases, explained answers and targeted correction.
Chapters 26–30
Retrieve after delay, adapt the method and monitor independence.
Complete chapter index
See the task · Chapters 1–6
Build precise meaning · Chapters 7–12
Make the product · Chapters 13–18
Practise and repair · Chapters 19–25
Transfer and check · Chapters 26–30
CHAPTER 1 OF 30 · See the task
1. Why facts become mixed during comparison
Why facts become mixed during comparison defines one part of the two-card source comparison. The learner needs to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it, but the visible difficulty can have several causes. Remembering the topic but losing the source is not the same problem as copying a dramatic detail instead of the relevant one. The first may concern selection or meaning; the second may concern sequence, reference or retrieval. This chapter narrows the task before adding language. A small accurate distinction gives the teacher something teachable and gives the learner a next move that can be attempted without carrying the whole assignment at once.
Begin why facts become mixed during comparison with a two-minute sample. The working case is this: two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin. The transfer case is different: two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator. Ask for a short comparison with traceable facts, then stop at the first point where the listener can no longer reconstruct the intended meaning. Do not begin by counting errors. Mark what remains stable: a clear noun, a useful verb, a correctly placed comparison or a sensible uncertainty. The diagnostic question is not “How many words does the learner know?” It is “Which connection failed first when the words had to serve this task?”
Listen specifically for remembering the topic but losing the source. If it appears, ask the learner to point to the detail that prompted the wording. Then check for copying a dramatic detail instead of the relevant one by changing only one feature of the case. Keeping most of the material constant makes the contrast visible. A correct response after a full model but an inaccurate response after a small change suggests dependence on imitation. A correct response after the change suggests that meaning, not surface wording, is beginning to guide the choice.
The first useful vocabulary pair for why facts become mixed during comparison is fact / inference. The second is similar topic / comparable feature. Say both sides aloud, place each in a short sentence and ask what information would make one choice better than the other. The learner should be able to reject a near miss, not merely repeat the accepted word. That rejection reveals the boundary of meaning. It also protects against the common habit of treating close-looking words as interchangeable whenever a sentence sounds fluent.
A proportionate repair for why facts become mixed during comparison uses one cue. If the missing link is meaning, contrast two examples. If it is reference, name the person, object, source or time again. If it is retrieval, hide the word and offer its job as the cue. If it is transfer, change the setting while retaining the decision. Give the learner time to act before supplying the answer. The pause is part of the diagnosis because immediate rescue can make assisted performance look like independent control.
Record the result of why facts become mixed during comparison in three columns: what the learner did accurately, the earliest unstable link, and the smallest next experiment. For the present chapter, the experiment may require source, however or habitat; it should not require all available vocabulary at once. Repeat the experiment later with a fresh surface case. If the same link fails again, teach it directly. If a different link fails, update the diagnosis rather than defending the first label.
Progress in why facts become mixed during comparison is visible when the learner needs less external language to begin, preserves the intended relation and can explain one important choice. It is not established by one polished performance. Check the same operation after a delay and in another mode, such as speaking before writing. This evidence remains local: it supports a teaching decision about the present task, not a general judgement about ability, maturity or the learner’s place in another education system.
CHAPTER 2 OF 30 · See the task
2. The Stage 3 reader job
The Stage 3 reader job defines one part of the two-card source comparison. The learner needs to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it, but the visible difficulty can have several causes. Comparing different categories as if they matched is not the same problem as using both when only one card supports the point. The first may concern selection or meaning; the second may concern sequence, reference or retrieval. This chapter narrows the task before adding language. A small accurate distinction gives the teacher something teachable and gives the learner a next move that can be attempted without carrying the whole assignment at once.
Begin the stage 3 reader job with a two-minute sample. The working case is this: two transport cards give journey distance, usual duration, number of stops and accessibility information. The transfer case is different: two weather cards report temperature range, rainfall observation, wind description and the date of recording. Ask for a short comparison with traceable facts, then stop at the first point where the listener can no longer reconstruct the intended meaning. Do not begin by counting errors. Mark what remains stable: a clear noun, a useful verb, a correctly placed comparison or a sensible uncertainty. The diagnostic question is not “How many words does the learner know?” It is “Which connection failed first when the words had to serve this task?”
Listen specifically for comparing different categories as if they matched. If it appears, ask the learner to point to the detail that prompted the wording. Then check for using both when only one card supports the point by changing only one feature of the case. Keeping most of the material constant makes the contrast visible. A correct response after a full model but an inaccurate response after a small change suggests dependence on imitation. A correct response after the change suggests that meaning, not surface wording, is beginning to guide the choice.
The first useful vocabulary pair for the stage 3 reader job is both / each. The second is whereas / however. Say both sides aloud, place each in a short sentence and ask what information would make one choice better than the other. The learner should be able to reject a near miss, not merely repeat the accepted word. That rejection reveals the boundary of meaning. It also protects against the common habit of treating close-looking words as interchangeable whenever a sentence sounds fluent.
A proportionate repair for the stage 3 reader job uses one cue. If the missing link is meaning, contrast two examples. If it is reference, name the person, object, source or time again. If it is retrieval, hide the word and offer its job as the cue. If it is transfer, change the setting while retaining the decision. Give the learner time to act before supplying the answer. The pause is part of the diagnosis because immediate rescue can make assisted performance look like independent control.
Record the result of the stage 3 reader job in three columns: what the learner did accurately, the earliest unstable link, and the smallest next experiment. For the present chapter, the experiment may require states, similar or distance; it should not require all available vocabulary at once. Repeat the experiment later with a fresh surface case. If the same link fails again, teach it directly. If a different link fails, update the diagnosis rather than defending the first label.
Progress in the stage 3 reader job is visible when the learner needs less external language to begin, preserves the intended relation and can explain one important choice. It is not established by one polished performance. Check the same operation after a delay and in another mode, such as speaking before writing. This evidence remains local: it supports a teaching decision about the present task, not a general judgement about ability, maturity or the learner’s place in another education system.
CHAPTER 3 OF 30 · See the task
3. Diagnose memory, source or comparison failure
Diagnose memory, source or comparison failure defines one part of the two-card source comparison. The learner needs to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it, but the visible difficulty can have several causes. Copying a dramatic detail instead of the relevant one is not the same problem as confusing not stated with false. The first may concern selection or meaning; the second may concern sequence, reference or retrieval. This chapter narrows the task before adding language. A small accurate distinction gives the teacher something teachable and gives the learner a next move that can be attempted without carrying the whole assignment at once.
Begin diagnose memory, source or comparison failure with a two-minute sample. The working case is this: two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator. The transfer case is different: two library-event cards list audience, start time, duration, materials to bring and registration conditions. Ask for a short comparison with traceable facts, then stop at the first point where the listener can no longer reconstruct the intended meaning. Do not begin by counting errors. Mark what remains stable: a clear noun, a useful verb, a correctly placed comparison or a sensible uncertainty. The diagnostic question is not “How many words does the learner know?” It is “Which connection failed first when the words had to serve this task?”
Listen specifically for copying a dramatic detail instead of the relevant one. If it appears, ask the learner to point to the detail that prompted the wording. Then check for confusing not stated with false by changing only one feature of the case. Keeping most of the material constant makes the contrast visible. A correct response after a full model but an inaccurate response after a small change suggests dependence on imitation. A correct response after the change suggests that meaning, not surface wording, is beginning to guide the choice.
The first useful vocabulary pair for diagnose memory, source or comparison failure is not stated / false. The second is interesting / relevant. Say both sides aloud, place each in a short sentence and ask what information would make one choice better than the other. The learner should be able to reject a near miss, not merely repeat the accepted word. That rejection reveals the boundary of meaning. It also protects against the common habit of treating close-looking words as interchangeable whenever a sentence sounds fluent.
A proportionate repair for diagnose memory, source or comparison failure uses one cue. If the missing link is meaning, contrast two examples. If it is reference, name the person, object, source or time again. If it is retrieval, hide the word and offer its job as the cue. If it is transfer, change the setting while retaining the decision. Give the learner time to act before supplying the answer. The pause is part of the diagnosis because immediate rescue can make assisted performance look like independent control.
Record the result of diagnose memory, source or comparison failure in three columns: what the learner did accurately, the earliest unstable link, and the smallest next experiment. For the present chapter, the experiment may require shows, different or duration; it should not require all available vocabulary at once. Repeat the experiment later with a fresh surface case. If the same link fails again, teach it directly. If a different link fails, update the diagnosis rather than defending the first label.
Progress in diagnose memory, source or comparison failure is visible when the learner needs less external language to begin, preserves the intended relation and can explain one important choice. It is not established by one polished performance. Check the same operation after a delay and in another mode, such as speaking before writing. This evidence remains local: it supports a teaching decision about the present task, not a general judgement about ability, maturity or the learner’s place in another education system.
CHAPTER 4 OF 30 · See the task
4. Give each card a stable name
Give each card a stable name defines one part of the two-card source comparison. The learner needs to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it, but the visible difficulty can have several causes. Using both when only one card supports the point is not the same problem as ending with a preference rather than a supported comparison. The first may concern selection or meaning; the second may concern sequence, reference or retrieval. This chapter narrows the task before adding language. A small accurate distinction gives the teacher something teachable and gives the learner a next move that can be attempted without carrying the whole assignment at once.
Begin give each card a stable name with a two-minute sample. The working case is this: two weather cards report temperature range, rainfall observation, wind description and the date of recording. The transfer case is different: two plant cards describe leaf edge, stem, growing place and whether flowering was observed. Ask for a short comparison with traceable facts, then stop at the first point where the listener can no longer reconstruct the intended meaning. Do not begin by counting errors. Mark what remains stable: a clear noun, a useful verb, a correctly placed comparison or a sensible uncertainty. The diagnostic question is not “How many words does the learner know?” It is “Which connection failed first when the words had to serve this task?”
Listen specifically for using both when only one card supports the point. If it appears, ask the learner to point to the detail that prompted the wording. Then check for ending with a preference rather than a supported comparison by changing only one feature of the case. Keeping most of the material constant makes the contrast visible. A correct response after a full model but an inaccurate response after a small change suggests dependence on imitation. A correct response after the change suggests that meaning, not surface wording, is beginning to guide the choice.
The first useful vocabulary pair for give each card a stable name is similar topic / comparable feature. The second is fact / inference. Say both sides aloud, place each in a short sentence and ask what information would make one choice better than the other. The learner should be able to reject a near miss, not merely repeat the accepted word. That rejection reveals the boundary of meaning. It also protects against the common habit of treating close-looking words as interchangeable whenever a sentence sounds fluent.
A proportionate repair for give each card a stable name uses one cue. If the missing link is meaning, contrast two examples. If it is reference, name the person, object, source or time again. If it is retrieval, hide the word and offer its job as the cue. If it is transfer, change the setting while retaining the decision. Give the learner time to act before supplying the answer. The pause is part of the diagnosis because immediate rescue can make assisted performance look like independent control.
Record the result of give each card a stable name in three columns: what the learner did accurately, the earliest unstable link, and the smallest next experiment. For the present chapter, the experiment may require both, feature or evidence; it should not require all available vocabulary at once. Repeat the experiment later with a fresh surface case. If the same link fails again, teach it directly. If a different link fails, update the diagnosis rather than defending the first label.
Progress in give each card a stable name is visible when the learner needs less external language to begin, preserves the intended relation and can explain one important choice. It is not established by one polished performance. Check the same operation after a delay and in another mode, such as speaking before writing. This evidence remains local: it supports a teaching decision about the present task, not a general judgement about ability, maturity or the learner’s place in another education system.
CHAPTER 5 OF 30 · See the task
5. Mark categories before details
Mark categories before details defines one part of the two-card source comparison. The learner needs to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it, but the visible difficulty can have several causes. Confusing not stated with false is not the same problem as remembering the topic but losing the source. The first may concern selection or meaning; the second may concern sequence, reference or retrieval. This chapter narrows the task before adding language. A small accurate distinction gives the teacher something teachable and gives the learner a next move that can be attempted without carrying the whole assignment at once.
Begin mark categories before details with a two-minute sample. The working case is this: two library-event cards list audience, start time, duration, materials to bring and registration conditions. The transfer case is different: two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin. Ask for a short comparison with traceable facts, then stop at the first point where the listener can no longer reconstruct the intended meaning. Do not begin by counting errors. Mark what remains stable: a clear noun, a useful verb, a correctly placed comparison or a sensible uncertainty. The diagnostic question is not “How many words does the learner know?” It is “Which connection failed first when the words had to serve this task?”
Listen specifically for confusing not stated with false. If it appears, ask the learner to point to the detail that prompted the wording. Then check for remembering the topic but losing the source by changing only one feature of the case. Keeping most of the material constant makes the contrast visible. A correct response after a full model but an inaccurate response after a small change suggests dependence on imitation. A correct response after the change suggests that meaning, not surface wording, is beginning to guide the choice.
The first useful vocabulary pair for mark categories before details is whereas / however. The second is both / each. Say both sides aloud, place each in a short sentence and ask what information would make one choice better than the other. The learner should be able to reject a near miss, not merely repeat the accepted word. That rejection reveals the boundary of meaning. It also protects against the common habit of treating close-looking words as interchangeable whenever a sentence sounds fluent.
A proportionate repair for mark categories before details uses one cue. If the missing link is meaning, contrast two examples. If it is reference, name the person, object, source or time again. If it is retrieval, hide the word and offer its job as the cue. If it is transfer, change the setting while retaining the decision. Give the learner time to act before supplying the answer. The pause is part of the diagnosis because immediate rescue can make assisted performance look like independent control.
Record the result of mark categories before details in three columns: what the learner did accurately, the earliest unstable link, and the smallest next experiment. For the present chapter, the experiment may require whereas, habitat or according to; it should not require all available vocabulary at once. Repeat the experiment later with a fresh surface case. If the same link fails again, teach it directly. If a different link fails, update the diagnosis rather than defending the first label.
Progress in mark categories before details is visible when the learner needs less external language to begin, preserves the intended relation and can explain one important choice. It is not established by one polished performance. Check the same operation after a delay and in another mode, such as speaking before writing. This evidence remains local: it supports a teaching decision about the present task, not a general judgement about ability, maturity or the learner’s place in another education system.
CHAPTER 6 OF 30 · See the task
6. Read one category across both cards
Read one category across both cards defines one part of the two-card source comparison. The learner needs to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it, but the visible difficulty can have several causes. Ending with a preference rather than a supported comparison is not the same problem as comparing different categories as if they matched. The first may concern selection or meaning; the second may concern sequence, reference or retrieval. This chapter narrows the task before adding language. A small accurate distinction gives the teacher something teachable and gives the learner a next move that can be attempted without carrying the whole assignment at once.
Begin read one category across both cards with a two-minute sample. The working case is this: two plant cards describe leaf edge, stem, growing place and whether flowering was observed. The transfer case is different: two transport cards give journey distance, usual duration, number of stops and accessibility information. Ask for a short comparison with traceable facts, then stop at the first point where the listener can no longer reconstruct the intended meaning. Do not begin by counting errors. Mark what remains stable: a clear noun, a useful verb, a correctly placed comparison or a sensible uncertainty. The diagnostic question is not “How many words does the learner know?” It is “Which connection failed first when the words had to serve this task?”
Listen specifically for ending with a preference rather than a supported comparison. If it appears, ask the learner to point to the detail that prompted the wording. Then check for comparing different categories as if they matched by changing only one feature of the case. Keeping most of the material constant makes the contrast visible. A correct response after a full model but an inaccurate response after a small change suggests dependence on imitation. A correct response after the change suggests that meaning, not surface wording, is beginning to guide the choice.
The first useful vocabulary pair for read one category across both cards is interesting / relevant. The second is not stated / false. Say both sides aloud, place each in a short sentence and ask what information would make one choice better than the other. The learner should be able to reject a near miss, not merely repeat the accepted word. That rejection reveals the boundary of meaning. It also protects against the common habit of treating close-looking words as interchangeable whenever a sentence sounds fluent.
A proportionate repair for read one category across both cards uses one cue. If the missing link is meaning, contrast two examples. If it is reference, name the person, object, source or time again. If it is retrieval, hide the word and offer its job as the cue. If it is transfer, change the setting while retaining the decision. Give the learner time to act before supplying the answer. The pause is part of the diagnosis because immediate rescue can make assisted performance look like independent control.
Record the result of read one category across both cards in three columns: what the learner did accurately, the earliest unstable link, and the smallest next experiment. For the present chapter, the experiment may require however, distance or not stated; it should not require all available vocabulary at once. Repeat the experiment later with a fresh surface case. If the same link fails again, teach it directly. If a different link fails, update the diagnosis rather than defending the first label.
Progress in read one category across both cards is visible when the learner needs less external language to begin, preserves the intended relation and can explain one important choice. It is not established by one polished performance. Check the same operation after a delay and in another mode, such as speaking before writing. This evidence remains local: it supports a teaching decision about the present task, not a general judgement about ability, maturity or the learner’s place in another education system.
CHAPTER 7 OF 30 · Build precise meaning
7. Use a two-column fact ledger
Use a two-column fact ledger is meaning work. In this chapter, similar, duration and conclusion are not offered as ornaments; each must earn a clear job in the two-card source comparison. Start by asking what the word points to, what it excludes and what form the sentence needs around it. A child may recognise a word in a list yet still lack its boundary, grammar or usual partners. The aim is therefore controlled use: a smaller vocabulary that represents the case accurately is stronger than a larger vocabulary used by guesswork.
Use fact / inference as the first boundary test for use a two-column fact ledger. Write one example that clearly fits each side, then one borderline example. Ask the learner to sort, justify and revise. Follow with similar topic / comparable feature, but keep the sentences short enough that the comparison—not reading load—remains the difficulty under inspection. A justification can be simple: “I chose this word because the detail shows …”. The reason matters more than repeating a definition, because it connects the word to evidence in the current context.
The working case is this: two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin. The transfer case is different: two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator. In the first case, invite the learner to underline the detail that licenses similar. In the second, ask whether duration still fits or whether another term is needed. This controlled change teaches that vocabulary choices respond to situations. It also reveals overgeneralisation: a word learned successfully in one example may be copied into every later example even when the relation has changed. Correction should make that shift visible, not merely replace the final word.
Build a compact word card for use a two-column fact ledger: the target term, a plain-language meaning, one near contrast, one natural sentence frame and one “do not use when” warning. For conclusion, the warning may be more valuable than another synonym. Keep the card beside the learner for guided practice, then turn it face down. Retrieval begins when the job or situation calls the word back without displaying it. If the learner can explain the job but not recall the label, offer the first sound only after a genuine attempt.
Grammar and collocation belong inside use a two-column fact ledger. Ask what usually comes before and after similar; check whether duration names an action, quality, relation or result in this sentence; and read the whole line aloud. A technically related word can still sound unnatural or change the intended claim. Rather than correcting every surface feature at once, preserve the successful relation and repair the smallest grammatical frame that lets the word operate. Then have the learner produce a second sentence independently.
Practise discrimination before production. Show three candidate sentences for two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin: one accurate, one plausible but unsupported, and one grammatically correct but semantically wrong. The learner chooses, explains and repairs. Only then ask for an original line using one of similar, duration or conclusion. This sequence lowers the initial production load while keeping thought active. It also gives the teacher evidence about whether the problem lies in recognition, meaning boundaries, sentence construction or retrieval.
Close use a two-column fact ledger with a delayed check. Remove the card, change to two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator, and ask the learner to recover the contrast and build one new sentence. Mark accuracy, clarity of reference and degree of prompting. Do not mark success merely because the target word appeared. A successful transfer preserves the relation even if the learner chooses simpler wording. A failed transfer is useful information: return to the boundary or retrieval cue that disappeared, then test again without enlarging the word list.
CHAPTER 8 OF 30 · Build precise meaning
8. Keep copied wording inside its source
Keep copied wording inside its source is meaning work. In this chapter, different, evidence and source are not offered as ornaments; each must earn a clear job in the two-card source comparison. Start by asking what the word points to, what it excludes and what form the sentence needs around it. A child may recognise a word in a list yet still lack its boundary, grammar or usual partners. The aim is therefore controlled use: a smaller vocabulary that represents the case accurately is stronger than a larger vocabulary used by guesswork.
Use both / each as the first boundary test for keep copied wording inside its source. Write one example that clearly fits each side, then one borderline example. Ask the learner to sort, justify and revise. Follow with whereas / however, but keep the sentences short enough that the comparison—not reading load—remains the difficulty under inspection. A justification can be simple: “I chose this word because the detail shows …”. The reason matters more than repeating a definition, because it connects the word to evidence in the current context.
The working case is this: two transport cards give journey distance, usual duration, number of stops and accessibility information. The transfer case is different: two weather cards report temperature range, rainfall observation, wind description and the date of recording. In the first case, invite the learner to underline the detail that licenses different. In the second, ask whether evidence still fits or whether another term is needed. This controlled change teaches that vocabulary choices respond to situations. It also reveals overgeneralisation: a word learned successfully in one example may be copied into every later example even when the relation has changed. Correction should make that shift visible, not merely replace the final word.
Build a compact word card for keep copied wording inside its source: the target term, a plain-language meaning, one near contrast, one natural sentence frame and one “do not use when” warning. For source, the warning may be more valuable than another synonym. Keep the card beside the learner for guided practice, then turn it face down. Retrieval begins when the job or situation calls the word back without displaying it. If the learner can explain the job but not recall the label, offer the first sound only after a genuine attempt.
Grammar and collocation belong inside keep copied wording inside its source. Ask what usually comes before and after different; check whether evidence names an action, quality, relation or result in this sentence; and read the whole line aloud. A technically related word can still sound unnatural or change the intended claim. Rather than correcting every surface feature at once, preserve the successful relation and repair the smallest grammatical frame that lets the word operate. Then have the learner produce a second sentence independently.
Practise discrimination before production. Show three candidate sentences for two transport cards give journey distance, usual duration, number of stops and accessibility information: one accurate, one plausible but unsupported, and one grammatically correct but semantically wrong. The learner chooses, explains and repairs. Only then ask for an original line using one of different, evidence or source. This sequence lowers the initial production load while keeping thought active. It also gives the teacher evidence about whether the problem lies in recognition, meaning boundaries, sentence construction or retrieval.
Close keep copied wording inside its source with a delayed check. Remove the card, change to two weather cards report temperature range, rainfall observation, wind description and the date of recording, and ask the learner to recover the contrast and build one new sentence. Mark accuracy, clarity of reference and degree of prompting. Do not mark success merely because the target word appeared. A successful transfer preserves the relation even if the learner chooses simpler wording. A failed transfer is useful information: return to the boundary or retrieval cue that disappeared, then test again without enlarging the word list.
CHAPTER 9 OF 30 · Build precise meaning
9. Paraphrase without changing the fact
Paraphrase without changing the fact is meaning work. In this chapter, feature, according to and states are not offered as ornaments; each must earn a clear job in the two-card source comparison. Start by asking what the word points to, what it excludes and what form the sentence needs around it. A child may recognise a word in a list yet still lack its boundary, grammar or usual partners. The aim is therefore controlled use: a smaller vocabulary that represents the case accurately is stronger than a larger vocabulary used by guesswork.
Use not stated / false as the first boundary test for paraphrase without changing the fact. Write one example that clearly fits each side, then one borderline example. Ask the learner to sort, justify and revise. Follow with interesting / relevant, but keep the sentences short enough that the comparison—not reading load—remains the difficulty under inspection. A justification can be simple: “I chose this word because the detail shows …”. The reason matters more than repeating a definition, because it connects the word to evidence in the current context.
The working case is this: two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator. The transfer case is different: two library-event cards list audience, start time, duration, materials to bring and registration conditions. In the first case, invite the learner to underline the detail that licenses feature. In the second, ask whether according to still fits or whether another term is needed. This controlled change teaches that vocabulary choices respond to situations. It also reveals overgeneralisation: a word learned successfully in one example may be copied into every later example even when the relation has changed. Correction should make that shift visible, not merely replace the final word.
Build a compact word card for paraphrase without changing the fact: the target term, a plain-language meaning, one near contrast, one natural sentence frame and one “do not use when” warning. For states, the warning may be more valuable than another synonym. Keep the card beside the learner for guided practice, then turn it face down. Retrieval begins when the job or situation calls the word back without displaying it. If the learner can explain the job but not recall the label, offer the first sound only after a genuine attempt.
Grammar and collocation belong inside paraphrase without changing the fact. Ask what usually comes before and after feature; check whether according to names an action, quality, relation or result in this sentence; and read the whole line aloud. A technically related word can still sound unnatural or change the intended claim. Rather than correcting every surface feature at once, preserve the successful relation and repair the smallest grammatical frame that lets the word operate. Then have the learner produce a second sentence independently.
Practise discrimination before production. Show three candidate sentences for two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator: one accurate, one plausible but unsupported, and one grammatically correct but semantically wrong. The learner chooses, explains and repairs. Only then ask for an original line using one of feature, according to or states. This sequence lowers the initial production load while keeping thought active. It also gives the teacher evidence about whether the problem lies in recognition, meaning boundaries, sentence construction or retrieval.
Close paraphrase without changing the fact with a delayed check. Remove the card, change to two library-event cards list audience, start time, duration, materials to bring and registration conditions, and ask the learner to recover the contrast and build one new sentence. Mark accuracy, clarity of reference and degree of prompting. Do not mark success merely because the target word appeared. A successful transfer preserves the relation even if the learner chooses simpler wording. A failed transfer is useful information: return to the boundary or retrieval cue that disappeared, then test again without enlarging the word list.
CHAPTER 10 OF 30 · Build precise meaning
10. Use both only for shared evidence
Use both only for shared evidence is meaning work. In this chapter, habitat, not stated and shows are not offered as ornaments; each must earn a clear job in the two-card source comparison. Start by asking what the word points to, what it excludes and what form the sentence needs around it. A child may recognise a word in a list yet still lack its boundary, grammar or usual partners. The aim is therefore controlled use: a smaller vocabulary that represents the case accurately is stronger than a larger vocabulary used by guesswork.
Use similar topic / comparable feature as the first boundary test for use both only for shared evidence. Write one example that clearly fits each side, then one borderline example. Ask the learner to sort, justify and revise. Follow with fact / inference, but keep the sentences short enough that the comparison—not reading load—remains the difficulty under inspection. A justification can be simple: “I chose this word because the detail shows …”. The reason matters more than repeating a definition, because it connects the word to evidence in the current context.
The working case is this: two weather cards report temperature range, rainfall observation, wind description and the date of recording. The transfer case is different: two plant cards describe leaf edge, stem, growing place and whether flowering was observed. In the first case, invite the learner to underline the detail that licenses habitat. In the second, ask whether not stated still fits or whether another term is needed. This controlled change teaches that vocabulary choices respond to situations. It also reveals overgeneralisation: a word learned successfully in one example may be copied into every later example even when the relation has changed. Correction should make that shift visible, not merely replace the final word.
Build a compact word card for use both only for shared evidence: the target term, a plain-language meaning, one near contrast, one natural sentence frame and one “do not use when” warning. For shows, the warning may be more valuable than another synonym. Keep the card beside the learner for guided practice, then turn it face down. Retrieval begins when the job or situation calls the word back without displaying it. If the learner can explain the job but not recall the label, offer the first sound only after a genuine attempt.
Grammar and collocation belong inside use both only for shared evidence. Ask what usually comes before and after habitat; check whether not stated names an action, quality, relation or result in this sentence; and read the whole line aloud. A technically related word can still sound unnatural or change the intended claim. Rather than correcting every surface feature at once, preserve the successful relation and repair the smallest grammatical frame that lets the word operate. Then have the learner produce a second sentence independently.
Practise discrimination before production. Show three candidate sentences for two weather cards report temperature range, rainfall observation, wind description and the date of recording: one accurate, one plausible but unsupported, and one grammatically correct but semantically wrong. The learner chooses, explains and repairs. Only then ask for an original line using one of habitat, not stated or shows. This sequence lowers the initial production load while keeping thought active. It also gives the teacher evidence about whether the problem lies in recognition, meaning boundaries, sentence construction or retrieval.
Close use both only for shared evidence with a delayed check. Remove the card, change to two plant cards describe leaf edge, stem, growing place and whether flowering was observed, and ask the learner to recover the contrast and build one new sentence. Mark accuracy, clarity of reference and degree of prompting. Do not mark success merely because the target word appeared. A successful transfer preserves the relation even if the learner chooses simpler wording. A failed transfer is useful information: return to the boundary or retrieval cue that disappeared, then test again without enlarging the word list.
CHAPTER 11 OF 30 · Build precise meaning
11. Use whereas for a matched contrast
Use whereas for a matched contrast is meaning work. In this chapter, distance, conclusion and both are not offered as ornaments; each must earn a clear job in the two-card source comparison. Start by asking what the word points to, what it excludes and what form the sentence needs around it. A child may recognise a word in a list yet still lack its boundary, grammar or usual partners. The aim is therefore controlled use: a smaller vocabulary that represents the case accurately is stronger than a larger vocabulary used by guesswork.
Use whereas / however as the first boundary test for use whereas for a matched contrast. Write one example that clearly fits each side, then one borderline example. Ask the learner to sort, justify and revise. Follow with both / each, but keep the sentences short enough that the comparison—not reading load—remains the difficulty under inspection. A justification can be simple: “I chose this word because the detail shows …”. The reason matters more than repeating a definition, because it connects the word to evidence in the current context.
The working case is this: two library-event cards list audience, start time, duration, materials to bring and registration conditions. The transfer case is different: two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin. In the first case, invite the learner to underline the detail that licenses distance. In the second, ask whether conclusion still fits or whether another term is needed. This controlled change teaches that vocabulary choices respond to situations. It also reveals overgeneralisation: a word learned successfully in one example may be copied into every later example even when the relation has changed. Correction should make that shift visible, not merely replace the final word.
Build a compact word card for use whereas for a matched contrast: the target term, a plain-language meaning, one near contrast, one natural sentence frame and one “do not use when” warning. For both, the warning may be more valuable than another synonym. Keep the card beside the learner for guided practice, then turn it face down. Retrieval begins when the job or situation calls the word back without displaying it. If the learner can explain the job but not recall the label, offer the first sound only after a genuine attempt.
Grammar and collocation belong inside use whereas for a matched contrast. Ask what usually comes before and after distance; check whether conclusion names an action, quality, relation or result in this sentence; and read the whole line aloud. A technically related word can still sound unnatural or change the intended claim. Rather than correcting every surface feature at once, preserve the successful relation and repair the smallest grammatical frame that lets the word operate. Then have the learner produce a second sentence independently.
Practise discrimination before production. Show three candidate sentences for two library-event cards list audience, start time, duration, materials to bring and registration conditions: one accurate, one plausible but unsupported, and one grammatically correct but semantically wrong. The learner chooses, explains and repairs. Only then ask for an original line using one of distance, conclusion or both. This sequence lowers the initial production load while keeping thought active. It also gives the teacher evidence about whether the problem lies in recognition, meaning boundaries, sentence construction or retrieval.
Close use whereas for a matched contrast with a delayed check. Remove the card, change to two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin, and ask the learner to recover the contrast and build one new sentence. Mark accuracy, clarity of reference and degree of prompting. Do not mark success merely because the target word appeared. A successful transfer preserves the relation even if the learner chooses simpler wording. A failed transfer is useful information: return to the boundary or retrieval cue that disappeared, then test again without enlarging the word list.
CHAPTER 12 OF 30 · Build precise meaning
12. Use however without changing the subject
Use however without changing the subject is meaning work. In this chapter, duration, source and whereas are not offered as ornaments; each must earn a clear job in the two-card source comparison. Start by asking what the word points to, what it excludes and what form the sentence needs around it. A child may recognise a word in a list yet still lack its boundary, grammar or usual partners. The aim is therefore controlled use: a smaller vocabulary that represents the case accurately is stronger than a larger vocabulary used by guesswork.
Use interesting / relevant as the first boundary test for use however without changing the subject. Write one example that clearly fits each side, then one borderline example. Ask the learner to sort, justify and revise. Follow with not stated / false, but keep the sentences short enough that the comparison—not reading load—remains the difficulty under inspection. A justification can be simple: “I chose this word because the detail shows …”. The reason matters more than repeating a definition, because it connects the word to evidence in the current context.
The working case is this: two plant cards describe leaf edge, stem, growing place and whether flowering was observed. The transfer case is different: two transport cards give journey distance, usual duration, number of stops and accessibility information. In the first case, invite the learner to underline the detail that licenses duration. In the second, ask whether source still fits or whether another term is needed. This controlled change teaches that vocabulary choices respond to situations. It also reveals overgeneralisation: a word learned successfully in one example may be copied into every later example even when the relation has changed. Correction should make that shift visible, not merely replace the final word.
Build a compact word card for use however without changing the subject: the target term, a plain-language meaning, one near contrast, one natural sentence frame and one “do not use when” warning. For whereas, the warning may be more valuable than another synonym. Keep the card beside the learner for guided practice, then turn it face down. Retrieval begins when the job or situation calls the word back without displaying it. If the learner can explain the job but not recall the label, offer the first sound only after a genuine attempt.
Grammar and collocation belong inside use however without changing the subject. Ask what usually comes before and after duration; check whether source names an action, quality, relation or result in this sentence; and read the whole line aloud. A technically related word can still sound unnatural or change the intended claim. Rather than correcting every surface feature at once, preserve the successful relation and repair the smallest grammatical frame that lets the word operate. Then have the learner produce a second sentence independently.
Practise discrimination before production. Show three candidate sentences for two plant cards describe leaf edge, stem, growing place and whether flowering was observed: one accurate, one plausible but unsupported, and one grammatically correct but semantically wrong. The learner chooses, explains and repairs. Only then ask for an original line using one of duration, source or whereas. This sequence lowers the initial production load while keeping thought active. It also gives the teacher evidence about whether the problem lies in recognition, meaning boundaries, sentence construction or retrieval.
Close use however without changing the subject with a delayed check. Remove the card, change to two transport cards give journey distance, usual duration, number of stops and accessibility information, and ask the learner to recover the contrast and build one new sentence. Mark accuracy, clarity of reference and degree of prompting. Do not mark success merely because the target word appeared. A successful transfer preserves the relation even if the learner chooses simpler wording. A failed transfer is useful information: return to the boundary or retrieval cue that disappeared, then test again without enlarging the word list.
CHAPTER 13 OF 30 · Make the product
13. Treat not stated as an evidence boundary
Treat not stated as an evidence boundary moves vocabulary into the finished a short comparison with traceable facts. The planning question is practical: what must the reader or listener understand at this point? Select one main relation, choose the smallest supporting detail and decide which of evidence, states or however makes that relation explicit. This prevents the product from becoming a showcase of every word taught earlier. Vocabulary serves structure when it helps the audience follow order, source, comparison, cause, constraint or decision without having to guess.
Plan treat not stated as an evidence boundary with visible boxes rather than a complete model answer. One box holds the relevant evidence from two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin; one holds the language job; one holds a possible sentence frame; and one holds a checking question. The learner fills the boxes before drafting. Because the plan separates evidence from wording, it becomes easier to revise a sentence without losing the underlying idea. It also lets a teacher see whether difficulty begins during selection or only when the idea is turned into connected language.
Model one decision for treat not stated as an evidence boundary, not the whole response. Say: “I need the audience to notice this relation. evidence may fit, but I will test it against fact / inference. I will use states only if the case supports that extra meaning.” Then stop and ask the learner to make the next decision. A useful model reveals uncertainty and checking. It does not pretend that good writing arrives fully formed, and it does not leave the learner copying a polished paragraph whose choices remain invisible.
Rehearse the product aloud for treat not stated as an evidence boundary. The learner should hear whether the referent of however is clear, whether the order can be followed and whether a connector promises more than the next clause delivers. Oral rehearsal is not a script to memorise word for word. After the meaning is stable, ask for a written version with one deliberate change: a more exact noun, a clearer source marker or a better ordered sentence. That small adaptation checks whether the learner owns the relation rather than the recital.
Draft treat not stated as an evidence boundary in two passes. The meaning pass checks that every important statement can be traced to the case. The language pass checks grammar, collocation and precision. On the meaning pass, circle any addition that is not supported by two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin. On the language pass, test evidence, states and however against a simpler alternative. Keep the precise word only when it improves the reader’s understanding. This order prevents surface polishing from concealing a weak or invented relationship.
Review treat not stated as an evidence boundary from the audience’s position. Can another person identify who or what each sentence refers to? Can they recover the sequence, comparison or decision without seeing the plan? Ask the audience to paraphrase the response. If the paraphrase changes the intended meaning, repair the relevant sentence. This is a stronger check than asking whether the writing “sounds good”, because it tests transmission. Clear vocabulary is successful when the intended relation survives movement from writer to reader.
Transfer the production method to two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator. Keep the boxes but remove the earlier wording. A learner who can select new evidence, recover the language job and create a fresh a short comparison with traceable facts is showing growing independence. A learner who reproduces the old sentence despite the changed facts needs another contrast, not a harsher judgement. Record which supports were needed and fade only one at a time. The objective is reliable choice under manageable load, followed by a check the learner can eventually run alone.
CHAPTER 14 OF 30 · Make the product
14. Separate fact from a reasonable inference
Separate fact from a reasonable inference moves vocabulary into the finished a short comparison with traceable facts. The planning question is practical: what must the reader or listener understand at this point? Select one main relation, choose the smallest supporting detail and decide which of according to, shows or similar makes that relation explicit. This prevents the product from becoming a showcase of every word taught earlier. Vocabulary serves structure when it helps the audience follow order, source, comparison, cause, constraint or decision without having to guess.
Plan separate fact from a reasonable inference with visible boxes rather than a complete model answer. One box holds the relevant evidence from two transport cards give journey distance, usual duration, number of stops and accessibility information; one holds the language job; one holds a possible sentence frame; and one holds a checking question. The learner fills the boxes before drafting. Because the plan separates evidence from wording, it becomes easier to revise a sentence without losing the underlying idea. It also lets a teacher see whether difficulty begins during selection or only when the idea is turned into connected language.
Model one decision for separate fact from a reasonable inference, not the whole response. Say: “I need the audience to notice this relation. according to may fit, but I will test it against both / each. I will use shows only if the case supports that extra meaning.” Then stop and ask the learner to make the next decision. A useful model reveals uncertainty and checking. It does not pretend that good writing arrives fully formed, and it does not leave the learner copying a polished paragraph whose choices remain invisible.
Rehearse the product aloud for separate fact from a reasonable inference. The learner should hear whether the referent of similar is clear, whether the order can be followed and whether a connector promises more than the next clause delivers. Oral rehearsal is not a script to memorise word for word. After the meaning is stable, ask for a written version with one deliberate change: a more exact noun, a clearer source marker or a better ordered sentence. That small adaptation checks whether the learner owns the relation rather than the recital.
Draft separate fact from a reasonable inference in two passes. The meaning pass checks that every important statement can be traced to the case. The language pass checks grammar, collocation and precision. On the meaning pass, circle any addition that is not supported by two transport cards give journey distance, usual duration, number of stops and accessibility information. On the language pass, test according to, shows and similar against a simpler alternative. Keep the precise word only when it improves the reader’s understanding. This order prevents surface polishing from concealing a weak or invented relationship.
Review separate fact from a reasonable inference from the audience’s position. Can another person identify who or what each sentence refers to? Can they recover the sequence, comparison or decision without seeing the plan? Ask the audience to paraphrase the response. If the paraphrase changes the intended meaning, repair the relevant sentence. This is a stronger check than asking whether the writing “sounds good”, because it tests transmission. Clear vocabulary is successful when the intended relation survives movement from writer to reader.
Transfer the production method to two weather cards report temperature range, rainfall observation, wind description and the date of recording. Keep the boxes but remove the earlier wording. A learner who can select new evidence, recover the language job and create a fresh a short comparison with traceable facts is showing growing independence. A learner who reproduces the old sentence despite the changed facts needs another contrast, not a harsher judgement. Record which supports were needed and fade only one at a time. The objective is reliable choice under manageable load, followed by a check the learner can eventually run alone.
CHAPTER 15 OF 30 · Make the product
15. Select relevant facts for the question
Select relevant facts for the question moves vocabulary into the finished a short comparison with traceable facts. The planning question is practical: what must the reader or listener understand at this point? Select one main relation, choose the smallest supporting detail and decide which of not stated, both or different makes that relation explicit. This prevents the product from becoming a showcase of every word taught earlier. Vocabulary serves structure when it helps the audience follow order, source, comparison, cause, constraint or decision without having to guess.
Plan select relevant facts for the question with visible boxes rather than a complete model answer. One box holds the relevant evidence from two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator; one holds the language job; one holds a possible sentence frame; and one holds a checking question. The learner fills the boxes before drafting. Because the plan separates evidence from wording, it becomes easier to revise a sentence without losing the underlying idea. It also lets a teacher see whether difficulty begins during selection or only when the idea is turned into connected language.
Model one decision for select relevant facts for the question, not the whole response. Say: “I need the audience to notice this relation. not stated may fit, but I will test it against not stated / false. I will use both only if the case supports that extra meaning.” Then stop and ask the learner to make the next decision. A useful model reveals uncertainty and checking. It does not pretend that good writing arrives fully formed, and it does not leave the learner copying a polished paragraph whose choices remain invisible.
Rehearse the product aloud for select relevant facts for the question. The learner should hear whether the referent of different is clear, whether the order can be followed and whether a connector promises more than the next clause delivers. Oral rehearsal is not a script to memorise word for word. After the meaning is stable, ask for a written version with one deliberate change: a more exact noun, a clearer source marker or a better ordered sentence. That small adaptation checks whether the learner owns the relation rather than the recital.
Draft select relevant facts for the question in two passes. The meaning pass checks that every important statement can be traced to the case. The language pass checks grammar, collocation and precision. On the meaning pass, circle any addition that is not supported by two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator. On the language pass, test not stated, both and different against a simpler alternative. Keep the precise word only when it improves the reader’s understanding. This order prevents surface polishing from concealing a weak or invented relationship.
Review select relevant facts for the question from the audience’s position. Can another person identify who or what each sentence refers to? Can they recover the sequence, comparison or decision without seeing the plan? Ask the audience to paraphrase the response. If the paraphrase changes the intended meaning, repair the relevant sentence. This is a stronger check than asking whether the writing “sounds good”, because it tests transmission. Clear vocabulary is successful when the intended relation survives movement from writer to reader.
Transfer the production method to two library-event cards list audience, start time, duration, materials to bring and registration conditions. Keep the boxes but remove the earlier wording. A learner who can select new evidence, recover the language job and create a fresh a short comparison with traceable facts is showing growing independence. A learner who reproduces the old sentence despite the changed facts needs another contrast, not a harsher judgement. Record which supports were needed and fade only one at a time. The objective is reliable choice under manageable load, followed by a check the learner can eventually run alone.
CHAPTER 16 OF 30 · Make the product
16. Build a comparison sentence pair
Build a comparison sentence pair moves vocabulary into the finished a short comparison with traceable facts. The planning question is practical: what must the reader or listener understand at this point? Select one main relation, choose the smallest supporting detail and decide which of conclusion, whereas or feature makes that relation explicit. This prevents the product from becoming a showcase of every word taught earlier. Vocabulary serves structure when it helps the audience follow order, source, comparison, cause, constraint or decision without having to guess.
Plan build a comparison sentence pair with visible boxes rather than a complete model answer. One box holds the relevant evidence from two weather cards report temperature range, rainfall observation, wind description and the date of recording; one holds the language job; one holds a possible sentence frame; and one holds a checking question. The learner fills the boxes before drafting. Because the plan separates evidence from wording, it becomes easier to revise a sentence without losing the underlying idea. It also lets a teacher see whether difficulty begins during selection or only when the idea is turned into connected language.
Model one decision for build a comparison sentence pair, not the whole response. Say: “I need the audience to notice this relation. conclusion may fit, but I will test it against similar topic / comparable feature. I will use whereas only if the case supports that extra meaning.” Then stop and ask the learner to make the next decision. A useful model reveals uncertainty and checking. It does not pretend that good writing arrives fully formed, and it does not leave the learner copying a polished paragraph whose choices remain invisible.
Rehearse the product aloud for build a comparison sentence pair. The learner should hear whether the referent of feature is clear, whether the order can be followed and whether a connector promises more than the next clause delivers. Oral rehearsal is not a script to memorise word for word. After the meaning is stable, ask for a written version with one deliberate change: a more exact noun, a clearer source marker or a better ordered sentence. That small adaptation checks whether the learner owns the relation rather than the recital.
Draft build a comparison sentence pair in two passes. The meaning pass checks that every important statement can be traced to the case. The language pass checks grammar, collocation and precision. On the meaning pass, circle any addition that is not supported by two weather cards report temperature range, rainfall observation, wind description and the date of recording. On the language pass, test conclusion, whereas and feature against a simpler alternative. Keep the precise word only when it improves the reader’s understanding. This order prevents surface polishing from concealing a weak or invented relationship.
Review build a comparison sentence pair from the audience’s position. Can another person identify who or what each sentence refers to? Can they recover the sequence, comparison or decision without seeing the plan? Ask the audience to paraphrase the response. If the paraphrase changes the intended meaning, repair the relevant sentence. This is a stronger check than asking whether the writing “sounds good”, because it tests transmission. Clear vocabulary is successful when the intended relation survives movement from writer to reader.
Transfer the production method to two plant cards describe leaf edge, stem, growing place and whether flowering was observed. Keep the boxes but remove the earlier wording. A learner who can select new evidence, recover the language job and create a fresh a short comparison with traceable facts is showing growing independence. A learner who reproduces the old sentence despite the changed facts needs another contrast, not a harsher judgement. Record which supports were needed and fade only one at a time. The objective is reliable choice under manageable load, followed by a check the learner can eventually run alone.
CHAPTER 17 OF 30 · Make the product
17. Write a conclusion smaller than the evidence
Write a conclusion smaller than the evidence moves vocabulary into the finished a short comparison with traceable facts. The planning question is practical: what must the reader or listener understand at this point? Select one main relation, choose the smallest supporting detail and decide which of source, however or habitat makes that relation explicit. This prevents the product from becoming a showcase of every word taught earlier. Vocabulary serves structure when it helps the audience follow order, source, comparison, cause, constraint or decision without having to guess.
Plan write a conclusion smaller than the evidence with visible boxes rather than a complete model answer. One box holds the relevant evidence from two library-event cards list audience, start time, duration, materials to bring and registration conditions; one holds the language job; one holds a possible sentence frame; and one holds a checking question. The learner fills the boxes before drafting. Because the plan separates evidence from wording, it becomes easier to revise a sentence without losing the underlying idea. It also lets a teacher see whether difficulty begins during selection or only when the idea is turned into connected language.
Model one decision for write a conclusion smaller than the evidence, not the whole response. Say: “I need the audience to notice this relation. source may fit, but I will test it against whereas / however. I will use however only if the case supports that extra meaning.” Then stop and ask the learner to make the next decision. A useful model reveals uncertainty and checking. It does not pretend that good writing arrives fully formed, and it does not leave the learner copying a polished paragraph whose choices remain invisible.
Rehearse the product aloud for write a conclusion smaller than the evidence. The learner should hear whether the referent of habitat is clear, whether the order can be followed and whether a connector promises more than the next clause delivers. Oral rehearsal is not a script to memorise word for word. After the meaning is stable, ask for a written version with one deliberate change: a more exact noun, a clearer source marker or a better ordered sentence. That small adaptation checks whether the learner owns the relation rather than the recital.
Draft write a conclusion smaller than the evidence in two passes. The meaning pass checks that every important statement can be traced to the case. The language pass checks grammar, collocation and precision. On the meaning pass, circle any addition that is not supported by two library-event cards list audience, start time, duration, materials to bring and registration conditions. On the language pass, test source, however and habitat against a simpler alternative. Keep the precise word only when it improves the reader’s understanding. This order prevents surface polishing from concealing a weak or invented relationship.
Review write a conclusion smaller than the evidence from the audience’s position. Can another person identify who or what each sentence refers to? Can they recover the sequence, comparison or decision without seeing the plan? Ask the audience to paraphrase the response. If the paraphrase changes the intended meaning, repair the relevant sentence. This is a stronger check than asking whether the writing “sounds good”, because it tests transmission. Clear vocabulary is successful when the intended relation survives movement from writer to reader.
Transfer the production method to two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin. Keep the boxes but remove the earlier wording. A learner who can select new evidence, recover the language job and create a fresh a short comparison with traceable facts is showing growing independence. A learner who reproduces the old sentence despite the changed facts needs another contrast, not a harsher judgement. Record which supports were needed and fade only one at a time. The objective is reliable choice under manageable load, followed by a check the learner can eventually run alone.
CHAPTER 18 OF 30 · Make the product
18. Add source language without sounding mechanical
Add source language without sounding mechanical moves vocabulary into the finished a short comparison with traceable facts. The planning question is practical: what must the reader or listener understand at this point? Select one main relation, choose the smallest supporting detail and decide which of states, similar or distance makes that relation explicit. This prevents the product from becoming a showcase of every word taught earlier. Vocabulary serves structure when it helps the audience follow order, source, comparison, cause, constraint or decision without having to guess.
Plan add source language without sounding mechanical with visible boxes rather than a complete model answer. One box holds the relevant evidence from two plant cards describe leaf edge, stem, growing place and whether flowering was observed; one holds the language job; one holds a possible sentence frame; and one holds a checking question. The learner fills the boxes before drafting. Because the plan separates evidence from wording, it becomes easier to revise a sentence without losing the underlying idea. It also lets a teacher see whether difficulty begins during selection or only when the idea is turned into connected language.
Model one decision for add source language without sounding mechanical, not the whole response. Say: “I need the audience to notice this relation. states may fit, but I will test it against interesting / relevant. I will use similar only if the case supports that extra meaning.” Then stop and ask the learner to make the next decision. A useful model reveals uncertainty and checking. It does not pretend that good writing arrives fully formed, and it does not leave the learner copying a polished paragraph whose choices remain invisible.
Rehearse the product aloud for add source language without sounding mechanical. The learner should hear whether the referent of distance is clear, whether the order can be followed and whether a connector promises more than the next clause delivers. Oral rehearsal is not a script to memorise word for word. After the meaning is stable, ask for a written version with one deliberate change: a more exact noun, a clearer source marker or a better ordered sentence. That small adaptation checks whether the learner owns the relation rather than the recital.
Draft add source language without sounding mechanical in two passes. The meaning pass checks that every important statement can be traced to the case. The language pass checks grammar, collocation and precision. On the meaning pass, circle any addition that is not supported by two plant cards describe leaf edge, stem, growing place and whether flowering was observed. On the language pass, test states, similar and distance against a simpler alternative. Keep the precise word only when it improves the reader’s understanding. This order prevents surface polishing from concealing a weak or invented relationship.
Review add source language without sounding mechanical from the audience’s position. Can another person identify who or what each sentence refers to? Can they recover the sequence, comparison or decision without seeing the plan? Ask the audience to paraphrase the response. If the paraphrase changes the intended meaning, repair the relevant sentence. This is a stronger check than asking whether the writing “sounds good”, because it tests transmission. Clear vocabulary is successful when the intended relation survives movement from writer to reader.
Transfer the production method to two transport cards give journey distance, usual duration, number of stops and accessibility information. Keep the boxes but remove the earlier wording. A learner who can select new evidence, recover the language job and create a fresh a short comparison with traceable facts is showing growing independence. A learner who reproduces the old sentence despite the changed facts needs another contrast, not a harsher judgement. Record which supports were needed and fade only one at a time. The objective is reliable choice under manageable load, followed by a check the learner can eventually run alone.
Practice, repair and explained answers
CHAPTER 19 OF 30 · Practise and repair
19. Repair a mixed-source sentence
Repair a mixed-source sentence is an original practice chapter built around two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin. Before reading any explanation, the learner should attempt a short comparison with traceable facts. Limit the first attempt so that the target relation remains visible: one short note, one comparison pair or one recommendation unit, depending on the task. The attempt creates evidence. Without it, an answer key can produce recognition while hiding whether the learner could retrieve shows, distinguish fact / inference or keep the facts attached to the right part of the case.
Question one for repair a mixed-source sentence: which detail matters most to the stated task, and why? Question two: which wording would overstate that detail? Question three: choose between shows and different, then explain the boundary. Question four: add duration only if it improves the relationship. These questions deliberately separate selection, meaning, certainty and sentence construction. If the learner misses one, keep the other successful steps visible. The repair should target the failed operation instead of restarting the entire exercise.
One defensible approach to repair a mixed-source sentence begins by naming the evidence plainly. It then uses a precise word to connect that evidence to the required job, while leaving unsupported information outside the sentence. For two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin, the explanation should make clear which detail comes from the material and which statement is an interpretation or decision. Another answer may choose different wording and still be correct. Accept alternatives when their meanings, references and degrees of certainty fit the case and can be justified.
A tempting error in repair a mixed-source sentence would be remembering the topic but losing the source. It may sound fluent because the topic is still recognisable, but the required relation has weakened. A second error would be copying a dramatic detail instead of the relevant one; this often appears when the learner remembers a word or detail without remembering its role. Place the error beside a repaired version and underline only the changed part. Ask what the revision fixes. That explanation turns correction into a reusable principle rather than a one-off replacement supplied by the answer page.
Extend repair a mixed-source sentence by altering one fact. Use the transfer case—two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator—but change the order, source, comparison point or practical constraint. Ask whether the original sentence still works. If it does not, the learner must identify the exact word or relationship that needs revision. This is a boundary test. It protects against memorised output and shows that vocabulary is responsive: a small change in evidence can require a different verb, connector, qualifier or conclusion.
Retrieve the lesson from repair a mixed-source sentence later without showing the completed answer. Offer the case title and one cue describing the language job. The learner reconstructs the relevant distinction, attempts a new sentence and then compares it with the earlier work. A delay may reveal that recognition was stronger than recall. That is not failure; it tells the teacher where practice belongs. Use a smaller cue next time only after the larger cue succeeds, and record the level of support accurately.
Mark repair a mixed-source sentence with a short rubric: evidence is traceable, the word meaning fits, reference is clear, grammar supports the meaning, and the learner can explain one choice. Do not award extra credit merely for length or rarity. If a simpler sentence meets all five checks, it is the stronger answer. If a sophisticated sentence fails the evidence or reference check, revise it. The chapter is complete when the learner understands both the answer and the method used to test it.
CHAPTER 20 OF 30 · Practise and repair
20. Repair an unmatched comparison
Repair an unmatched comparison is an original practice chapter built around two transport cards give journey distance, usual duration, number of stops and accessibility information. Before reading any explanation, the learner should attempt a short comparison with traceable facts. Limit the first attempt so that the target relation remains visible: one short note, one comparison pair or one recommendation unit, depending on the task. The attempt creates evidence. Without it, an answer key can produce recognition while hiding whether the learner could retrieve both, distinguish both / each or keep the facts attached to the right part of the case.
Question one for repair an unmatched comparison: which detail matters most to the stated task, and why? Question two: which wording would overstate that detail? Question three: choose between both and feature, then explain the boundary. Question four: add evidence only if it improves the relationship. These questions deliberately separate selection, meaning, certainty and sentence construction. If the learner misses one, keep the other successful steps visible. The repair should target the failed operation instead of restarting the entire exercise.
One defensible approach to repair an unmatched comparison begins by naming the evidence plainly. It then uses a precise word to connect that evidence to the required job, while leaving unsupported information outside the sentence. For two transport cards give journey distance, usual duration, number of stops and accessibility information, the explanation should make clear which detail comes from the material and which statement is an interpretation or decision. Another answer may choose different wording and still be correct. Accept alternatives when their meanings, references and degrees of certainty fit the case and can be justified.
A tempting error in repair an unmatched comparison would be comparing different categories as if they matched. It may sound fluent because the topic is still recognisable, but the required relation has weakened. A second error would be using both when only one card supports the point; this often appears when the learner remembers a word or detail without remembering its role. Place the error beside a repaired version and underline only the changed part. Ask what the revision fixes. That explanation turns correction into a reusable principle rather than a one-off replacement supplied by the answer page.
Extend repair an unmatched comparison by altering one fact. Use the transfer case—two weather cards report temperature range, rainfall observation, wind description and the date of recording—but change the order, source, comparison point or practical constraint. Ask whether the original sentence still works. If it does not, the learner must identify the exact word or relationship that needs revision. This is a boundary test. It protects against memorised output and shows that vocabulary is responsive: a small change in evidence can require a different verb, connector, qualifier or conclusion.
Retrieve the lesson from repair an unmatched comparison later without showing the completed answer. Offer the case title and one cue describing the language job. The learner reconstructs the relevant distinction, attempts a new sentence and then compares it with the earlier work. A delay may reveal that recognition was stronger than recall. That is not failure; it tells the teacher where practice belongs. Use a smaller cue next time only after the larger cue succeeds, and record the level of support accurately.
Mark repair an unmatched comparison with a short rubric: evidence is traceable, the word meaning fits, reference is clear, grammar supports the meaning, and the learner can explain one choice. Do not award extra credit merely for length or rarity. If a simpler sentence meets all five checks, it is the stronger answer. If a sophisticated sentence fails the evidence or reference check, revise it. The chapter is complete when the learner understands both the answer and the method used to test it.
CHAPTER 21 OF 30 · Practise and repair
21. Repair a conclusion based on preference
Repair a conclusion based on preference is an original practice chapter built around two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator. Before reading any explanation, the learner should attempt a short comparison with traceable facts. Limit the first attempt so that the target relation remains visible: one short note, one comparison pair or one recommendation unit, depending on the task. The attempt creates evidence. Without it, an answer key can produce recognition while hiding whether the learner could retrieve whereas, distinguish not stated / false or keep the facts attached to the right part of the case.
Question one for repair a conclusion based on preference: which detail matters most to the stated task, and why? Question two: which wording would overstate that detail? Question three: choose between whereas and habitat, then explain the boundary. Question four: add according to only if it improves the relationship. These questions deliberately separate selection, meaning, certainty and sentence construction. If the learner misses one, keep the other successful steps visible. The repair should target the failed operation instead of restarting the entire exercise.
One defensible approach to repair a conclusion based on preference begins by naming the evidence plainly. It then uses a precise word to connect that evidence to the required job, while leaving unsupported information outside the sentence. For two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator, the explanation should make clear which detail comes from the material and which statement is an interpretation or decision. Another answer may choose different wording and still be correct. Accept alternatives when their meanings, references and degrees of certainty fit the case and can be justified.
A tempting error in repair a conclusion based on preference would be copying a dramatic detail instead of the relevant one. It may sound fluent because the topic is still recognisable, but the required relation has weakened. A second error would be confusing not stated with false; this often appears when the learner remembers a word or detail without remembering its role. Place the error beside a repaired version and underline only the changed part. Ask what the revision fixes. That explanation turns correction into a reusable principle rather than a one-off replacement supplied by the answer page.
Extend repair a conclusion based on preference by altering one fact. Use the transfer case—two library-event cards list audience, start time, duration, materials to bring and registration conditions—but change the order, source, comparison point or practical constraint. Ask whether the original sentence still works. If it does not, the learner must identify the exact word or relationship that needs revision. This is a boundary test. It protects against memorised output and shows that vocabulary is responsive: a small change in evidence can require a different verb, connector, qualifier or conclusion.
Retrieve the lesson from repair a conclusion based on preference later without showing the completed answer. Offer the case title and one cue describing the language job. The learner reconstructs the relevant distinction, attempts a new sentence and then compares it with the earlier work. A delay may reveal that recognition was stronger than recall. That is not failure; it tells the teacher where practice belongs. Use a smaller cue next time only after the larger cue succeeds, and record the level of support accurately.
Mark repair a conclusion based on preference with a short rubric: evidence is traceable, the word meaning fits, reference is clear, grammar supports the meaning, and the learner can explain one choice. Do not award extra credit merely for length or rarity. If a simpler sentence meets all five checks, it is the stronger answer. If a sophisticated sentence fails the evidence or reference check, revise it. The chapter is complete when the learner understands both the answer and the method used to test it.
CHAPTER 22 OF 30 · Practise and repair
22. Practice set one: otter and pangolin
Practice set one: otter and pangolin is an original practice chapter built around two weather cards report temperature range, rainfall observation, wind description and the date of recording. Before reading any explanation, the learner should attempt a short comparison with traceable facts. Limit the first attempt so that the target relation remains visible: one short note, one comparison pair or one recommendation unit, depending on the task. The attempt creates evidence. Without it, an answer key can produce recognition while hiding whether the learner could retrieve however, distinguish similar topic / comparable feature or keep the facts attached to the right part of the case.
Question one for practice set one: otter and pangolin: which detail matters most to the stated task, and why? Question two: which wording would overstate that detail? Question three: choose between however and distance, then explain the boundary. Question four: add not stated only if it improves the relationship. These questions deliberately separate selection, meaning, certainty and sentence construction. If the learner misses one, keep the other successful steps visible. The repair should target the failed operation instead of restarting the entire exercise.
One defensible approach to practice set one: otter and pangolin begins by naming the evidence plainly. It then uses a precise word to connect that evidence to the required job, while leaving unsupported information outside the sentence. For two weather cards report temperature range, rainfall observation, wind description and the date of recording, the explanation should make clear which detail comes from the material and which statement is an interpretation or decision. Another answer may choose different wording and still be correct. Accept alternatives when their meanings, references and degrees of certainty fit the case and can be justified.
A tempting error in practice set one: otter and pangolin would be using both when only one card supports the point. It may sound fluent because the topic is still recognisable, but the required relation has weakened. A second error would be ending with a preference rather than a supported comparison; this often appears when the learner remembers a word or detail without remembering its role. Place the error beside a repaired version and underline only the changed part. Ask what the revision fixes. That explanation turns correction into a reusable principle rather than a one-off replacement supplied by the answer page.
Extend practice set one: otter and pangolin by altering one fact. Use the transfer case—two plant cards describe leaf edge, stem, growing place and whether flowering was observed—but change the order, source, comparison point or practical constraint. Ask whether the original sentence still works. If it does not, the learner must identify the exact word or relationship that needs revision. This is a boundary test. It protects against memorised output and shows that vocabulary is responsive: a small change in evidence can require a different verb, connector, qualifier or conclusion.
Retrieve the lesson from practice set one: otter and pangolin later without showing the completed answer. Offer the case title and one cue describing the language job. The learner reconstructs the relevant distinction, attempts a new sentence and then compares it with the earlier work. A delay may reveal that recognition was stronger than recall. That is not failure; it tells the teacher where practice belongs. Use a smaller cue next time only after the larger cue succeeds, and record the level of support accurately.
Mark practice set one: otter and pangolin with a short rubric: evidence is traceable, the word meaning fits, reference is clear, grammar supports the meaning, and the learner can explain one choice. Do not award extra credit merely for length or rarity. If a simpler sentence meets all five checks, it is the stronger answer. If a sophisticated sentence fails the evidence or reference check, revise it. The chapter is complete when the learner understands both the answer and the method used to test it.
CHAPTER 23 OF 30 · Practise and repair
23. Practice set two: two journeys
Practice set two: two journeys is an original practice chapter built around two library-event cards list audience, start time, duration, materials to bring and registration conditions. Before reading any explanation, the learner should attempt a short comparison with traceable facts. Limit the first attempt so that the target relation remains visible: one short note, one comparison pair or one recommendation unit, depending on the task. The attempt creates evidence. Without it, an answer key can produce recognition while hiding whether the learner could retrieve similar, distinguish whereas / however or keep the facts attached to the right part of the case.
Question one for practice set two: two journeys: which detail matters most to the stated task, and why? Question two: which wording would overstate that detail? Question three: choose between similar and duration, then explain the boundary. Question four: add conclusion only if it improves the relationship. These questions deliberately separate selection, meaning, certainty and sentence construction. If the learner misses one, keep the other successful steps visible. The repair should target the failed operation instead of restarting the entire exercise.
One defensible approach to practice set two: two journeys begins by naming the evidence plainly. It then uses a precise word to connect that evidence to the required job, while leaving unsupported information outside the sentence. For two library-event cards list audience, start time, duration, materials to bring and registration conditions, the explanation should make clear which detail comes from the material and which statement is an interpretation or decision. Another answer may choose different wording and still be correct. Accept alternatives when their meanings, references and degrees of certainty fit the case and can be justified.
A tempting error in practice set two: two journeys would be confusing not stated with false. It may sound fluent because the topic is still recognisable, but the required relation has weakened. A second error would be remembering the topic but losing the source; this often appears when the learner remembers a word or detail without remembering its role. Place the error beside a repaired version and underline only the changed part. Ask what the revision fixes. That explanation turns correction into a reusable principle rather than a one-off replacement supplied by the answer page.
Extend practice set two: two journeys by altering one fact. Use the transfer case—two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin—but change the order, source, comparison point or practical constraint. Ask whether the original sentence still works. If it does not, the learner must identify the exact word or relationship that needs revision. This is a boundary test. It protects against memorised output and shows that vocabulary is responsive: a small change in evidence can require a different verb, connector, qualifier or conclusion.
Retrieve the lesson from practice set two: two journeys later without showing the completed answer. Offer the case title and one cue describing the language job. The learner reconstructs the relevant distinction, attempts a new sentence and then compares it with the earlier work. A delay may reveal that recognition was stronger than recall. That is not failure; it tells the teacher where practice belongs. Use a smaller cue next time only after the larger cue succeeds, and record the level of support accurately.
Mark practice set two: two journeys with a short rubric: evidence is traceable, the word meaning fits, reference is clear, grammar supports the meaning, and the learner can explain one choice. Do not award extra credit merely for length or rarity. If a simpler sentence meets all five checks, it is the stronger answer. If a sophisticated sentence fails the evidence or reference check, revise it. The chapter is complete when the learner understands both the answer and the method used to test it.
CHAPTER 24 OF 30 · Practise and repair
24. Practice set three: two museum objects
Practice set three: two museum objects is an original practice chapter built around two plant cards describe leaf edge, stem, growing place and whether flowering was observed. Before reading any explanation, the learner should attempt a short comparison with traceable facts. Limit the first attempt so that the target relation remains visible: one short note, one comparison pair or one recommendation unit, depending on the task. The attempt creates evidence. Without it, an answer key can produce recognition while hiding whether the learner could retrieve different, distinguish interesting / relevant or keep the facts attached to the right part of the case.
Question one for practice set three: two museum objects: which detail matters most to the stated task, and why? Question two: which wording would overstate that detail? Question three: choose between different and evidence, then explain the boundary. Question four: add source only if it improves the relationship. These questions deliberately separate selection, meaning, certainty and sentence construction. If the learner misses one, keep the other successful steps visible. The repair should target the failed operation instead of restarting the entire exercise.
One defensible approach to practice set three: two museum objects begins by naming the evidence plainly. It then uses a precise word to connect that evidence to the required job, while leaving unsupported information outside the sentence. For two plant cards describe leaf edge, stem, growing place and whether flowering was observed, the explanation should make clear which detail comes from the material and which statement is an interpretation or decision. Another answer may choose different wording and still be correct. Accept alternatives when their meanings, references and degrees of certainty fit the case and can be justified.
A tempting error in practice set three: two museum objects would be ending with a preference rather than a supported comparison. It may sound fluent because the topic is still recognisable, but the required relation has weakened. A second error would be comparing different categories as if they matched; this often appears when the learner remembers a word or detail without remembering its role. Place the error beside a repaired version and underline only the changed part. Ask what the revision fixes. That explanation turns correction into a reusable principle rather than a one-off replacement supplied by the answer page.
Extend practice set three: two museum objects by altering one fact. Use the transfer case—two transport cards give journey distance, usual duration, number of stops and accessibility information—but change the order, source, comparison point or practical constraint. Ask whether the original sentence still works. If it does not, the learner must identify the exact word or relationship that needs revision. This is a boundary test. It protects against memorised output and shows that vocabulary is responsive: a small change in evidence can require a different verb, connector, qualifier or conclusion.
Retrieve the lesson from practice set three: two museum objects later without showing the completed answer. Offer the case title and one cue describing the language job. The learner reconstructs the relevant distinction, attempts a new sentence and then compares it with the earlier work. A delay may reveal that recognition was stronger than recall. That is not failure; it tells the teacher where practice belongs. Use a smaller cue next time only after the larger cue succeeds, and record the level of support accurately.
Mark practice set three: two museum objects with a short rubric: evidence is traceable, the word meaning fits, reference is clear, grammar supports the meaning, and the learner can explain one choice. Do not award extra credit merely for length or rarity. If a simpler sentence meets all five checks, it is the stronger answer. If a sophisticated sentence fails the evidence or reference check, revise it. The chapter is complete when the learner understands both the answer and the method used to test it.
CHAPTER 25 OF 30 · Practise and repair
25. Practice set four: two event notices
Practice set four: two event notices is an original practice chapter built around two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin. Before reading any explanation, the learner should attempt a short comparison with traceable facts. Limit the first attempt so that the target relation remains visible: one short note, one comparison pair or one recommendation unit, depending on the task. The attempt creates evidence. Without it, an answer key can produce recognition while hiding whether the learner could retrieve feature, distinguish fact / inference or keep the facts attached to the right part of the case.
Question one for practice set four: two event notices: which detail matters most to the stated task, and why? Question two: which wording would overstate that detail? Question three: choose between feature and according to, then explain the boundary. Question four: add states only if it improves the relationship. These questions deliberately separate selection, meaning, certainty and sentence construction. If the learner misses one, keep the other successful steps visible. The repair should target the failed operation instead of restarting the entire exercise.
One defensible approach to practice set four: two event notices begins by naming the evidence plainly. It then uses a precise word to connect that evidence to the required job, while leaving unsupported information outside the sentence. For two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin, the explanation should make clear which detail comes from the material and which statement is an interpretation or decision. Another answer may choose different wording and still be correct. Accept alternatives when their meanings, references and degrees of certainty fit the case and can be justified.
A tempting error in practice set four: two event notices would be remembering the topic but losing the source. It may sound fluent because the topic is still recognisable, but the required relation has weakened. A second error would be copying a dramatic detail instead of the relevant one; this often appears when the learner remembers a word or detail without remembering its role. Place the error beside a repaired version and underline only the changed part. Ask what the revision fixes. That explanation turns correction into a reusable principle rather than a one-off replacement supplied by the answer page.
Extend practice set four: two event notices by altering one fact. Use the transfer case—two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator—but change the order, source, comparison point or practical constraint. Ask whether the original sentence still works. If it does not, the learner must identify the exact word or relationship that needs revision. This is a boundary test. It protects against memorised output and shows that vocabulary is responsive: a small change in evidence can require a different verb, connector, qualifier or conclusion.
Retrieve the lesson from practice set four: two event notices later without showing the completed answer. Offer the case title and one cue describing the language job. The learner reconstructs the relevant distinction, attempts a new sentence and then compares it with the earlier work. A delay may reveal that recognition was stronger than recall. That is not failure; it tells the teacher where practice belongs. Use a smaller cue next time only after the larger cue succeeds, and record the level of support accurately.
Mark practice set four: two event notices with a short rubric: evidence is traceable, the word meaning fits, reference is clear, grammar supports the meaning, and the learner can explain one choice. Do not award extra credit merely for length or rarity. If a simpler sentence meets all five checks, it is the stronger answer. If a sophisticated sentence fails the evidence or reference check, revise it. The chapter is complete when the learner understands both the answer and the method used to test it.
CHAPTER 26 OF 30 · Transfer and check
26. Explained answers and why alternatives work
Explained answers and why alternatives work checks whether the language work has become portable. The original aim remains to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it. Portability means that the learner can recognise the job in a fresh situation, retrieve a workable term, adapt the sentence and check the result. It does not require identical speed or wording every time. Begin with two transport cards give journey distance, usual duration, number of stops and accessibility information, remove the completed model and note whether the learner asks for a word, a meaning distinction, a task reminder or no help at all.
Use a delayed retrieval check for explained answers and why alternatives work. Give the task cue rather than the target list: ask for the word that marks order, identifies a source, names a change, preserves a counterpoint or signals a recommendation, as appropriate. If habitat does not arrive, invite a plain-language explanation before offering a sound cue. Recovering the idea first keeps meaning in charge. After recall, the learner should use the word in a new sentence so that memory and application are checked together.
Change mode during explained answers and why alternatives work. Move from spoken explanation to a written line, from a card to a simple table, or from an individual response to a short partner discussion. Keep the underlying distinction both / each visible. A word that works only in the original format is not yet reliably transferable. If the switch creates too much load, restore one support—perhaps the evidence boxes—while leaving the vocabulary hidden. The aim is to isolate the new difficulty instead of treating every stumble as forgetting.
Fade support for explained answers and why alternatives work in an observable sequence: full worked example, incomplete frame, job cue, first-sound cue and independent attempt. Do not remove several supports together and then guess why performance fell. For two weather cards report temperature range, rainfall observation, wind description and the date of recording, record the lightest cue that produced an accurate response and whether the learner could explain the choice. Over time, the pattern should show which distinctions are secure, which need occasional retrieval and which still depend on adult language.
Families can support explained answers and why alternatives work with a short conversation, not a second classroom. Ask the child to explain one choice from a picture, notice, card or group decision. Teachers can preserve the same cue language across subjects. Learners can keep a tiny record of words that solved a real communication problem. None of these actions needs a daily score. The useful evidence is whether the child starts more readily, asks a more precise question, notices a mismatch or repairs a sentence with less help.
Use limits responsibly in explained answers and why alternatives work. Cambridge Primary Stage 3 is used as an official curriculum-stage label. This article does not convert it into a Singapore, British, American or other school level, and the cards are original teaching texts rather than Cambridge assessment materials. A child may show strong understanding through simpler words, or know a word but need time to retrieve it. Multilingual knowledge can support meaning even when English production is still developing. The exercises should therefore inform the next piece of teaching rather than serve as a universal maturity test. Follow the school’s language pathway and adapt examples so that the learner’s cultural knowledge is a resource, not an unspoken prerequisite.
The final progress check for explained answers and why alternatives work has four parts: accurate local meaning, clear reference, natural sentence fit and later independent use. Add a fifth only when relevant: the learner can explain why an alternative would be weaker. Keep the record descriptive—independent, small cue, full cue or not yet secure. Then choose one next experiment. Progress is the growing ability to use vocabulary for meaning, comprehension, writing and learning while carrying a sensible check of one’s own work.
CHAPTER 27 OF 30 · Transfer and check
27. Retrieve source markers after a delay
Retrieve source markers after a delay checks whether the language work has become portable. The original aim remains to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it. Portability means that the learner can recognise the job in a fresh situation, retrieve a workable term, adapt the sentence and check the result. It does not require identical speed or wording every time. Begin with two historical object cards identify material, known use, discovery place and an uncertainty noted by a curator, remove the completed model and note whether the learner asks for a word, a meaning distinction, a task reminder or no help at all.
Use a delayed retrieval check for retrieve source markers after a delay. Give the task cue rather than the target list: ask for the word that marks order, identifies a source, names a change, preserves a counterpoint or signals a recommendation, as appropriate. If distance does not arrive, invite a plain-language explanation before offering a sound cue. Recovering the idea first keeps meaning in charge. After recall, the learner should use the word in a new sentence so that memory and application are checked together.
Change mode during retrieve source markers after a delay. Move from spoken explanation to a written line, from a card to a simple table, or from an individual response to a short partner discussion. Keep the underlying distinction not stated / false visible. A word that works only in the original format is not yet reliably transferable. If the switch creates too much load, restore one support—perhaps the evidence boxes—while leaving the vocabulary hidden. The aim is to isolate the new difficulty instead of treating every stumble as forgetting.
Fade support for retrieve source markers after a delay in an observable sequence: full worked example, incomplete frame, job cue, first-sound cue and independent attempt. Do not remove several supports together and then guess why performance fell. For two library-event cards list audience, start time, duration, materials to bring and registration conditions, record the lightest cue that produced an accurate response and whether the learner could explain the choice. Over time, the pattern should show which distinctions are secure, which need occasional retrieval and which still depend on adult language.
Families can support retrieve source markers after a delay with a short conversation, not a second classroom. Ask the child to explain one choice from a picture, notice, card or group decision. Teachers can preserve the same cue language across subjects. Learners can keep a tiny record of words that solved a real communication problem. None of these actions needs a daily score. The useful evidence is whether the child starts more readily, asks a more precise question, notices a mismatch or repairs a sentence with less help.
Use limits responsibly in retrieve source markers after a delay. Cambridge Primary Stage 3 is used as an official curriculum-stage label. This article does not convert it into a Singapore, British, American or other school level, and the cards are original teaching texts rather than Cambridge assessment materials. A child may show strong understanding through simpler words, or know a word but need time to retrieve it. Multilingual knowledge can support meaning even when English production is still developing. The exercises should therefore inform the next piece of teaching rather than serve as a universal maturity test. Follow the school’s language pathway and adapt examples so that the learner’s cultural knowledge is a resource, not an unspoken prerequisite.
The final progress check for retrieve source markers after a delay has four parts: accurate local meaning, clear reference, natural sentence fit and later independent use. Add a fifth only when relevant: the learner can explain why an alternative would be weaker. Keep the record descriptive—independent, small cue, full cue or not yet secure. Then choose one next experiment. Progress is the growing ability to use vocabulary for meaning, comprehension, writing and learning while carrying a sensible check of one’s own work.
CHAPTER 28 OF 30 · Transfer and check
28. Transfer to web pages, diagrams and tables
Transfer to web pages, diagrams and tables checks whether the language work has become portable. The original aim remains to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it. Portability means that the learner can recognise the job in a fresh situation, retrieve a workable term, adapt the sentence and check the result. It does not require identical speed or wording every time. Begin with two weather cards report temperature range, rainfall observation, wind description and the date of recording, remove the completed model and note whether the learner asks for a word, a meaning distinction, a task reminder or no help at all.
Use a delayed retrieval check for transfer to web pages, diagrams and tables. Give the task cue rather than the target list: ask for the word that marks order, identifies a source, names a change, preserves a counterpoint or signals a recommendation, as appropriate. If duration does not arrive, invite a plain-language explanation before offering a sound cue. Recovering the idea first keeps meaning in charge. After recall, the learner should use the word in a new sentence so that memory and application are checked together.
Change mode during transfer to web pages, diagrams and tables. Move from spoken explanation to a written line, from a card to a simple table, or from an individual response to a short partner discussion. Keep the underlying distinction similar topic / comparable feature visible. A word that works only in the original format is not yet reliably transferable. If the switch creates too much load, restore one support—perhaps the evidence boxes—while leaving the vocabulary hidden. The aim is to isolate the new difficulty instead of treating every stumble as forgetting.
Fade support for transfer to web pages, diagrams and tables in an observable sequence: full worked example, incomplete frame, job cue, first-sound cue and independent attempt. Do not remove several supports together and then guess why performance fell. For two plant cards describe leaf edge, stem, growing place and whether flowering was observed, record the lightest cue that produced an accurate response and whether the learner could explain the choice. Over time, the pattern should show which distinctions are secure, which need occasional retrieval and which still depend on adult language.
Families can support transfer to web pages, diagrams and tables with a short conversation, not a second classroom. Ask the child to explain one choice from a picture, notice, card or group decision. Teachers can preserve the same cue language across subjects. Learners can keep a tiny record of words that solved a real communication problem. None of these actions needs a daily score. The useful evidence is whether the child starts more readily, asks a more precise question, notices a mismatch or repairs a sentence with less help.
Use limits responsibly in transfer to web pages, diagrams and tables. Cambridge Primary Stage 3 is used as an official curriculum-stage label. This article does not convert it into a Singapore, British, American or other school level, and the cards are original teaching texts rather than Cambridge assessment materials. A child may show strong understanding through simpler words, or know a word but need time to retrieve it. Multilingual knowledge can support meaning even when English production is still developing. The exercises should therefore inform the next piece of teaching rather than serve as a universal maturity test. Follow the school’s language pathway and adapt examples so that the learner’s cultural knowledge is a resource, not an unspoken prerequisite.
The final progress check for transfer to web pages, diagrams and tables has four parts: accurate local meaning, clear reference, natural sentence fit and later independent use. Add a fifth only when relevant: the learner can explain why an alternative would be weaker. Keep the record descriptive—independent, small cue, full cue or not yet secure. Then choose one next experiment. Progress is the growing ability to use vocabulary for meaning, comprehension, writing and learning while carrying a sensible check of one’s own work.
CHAPTER 29 OF 30 · Transfer and check
29. A four-week fact-tracking route
A four-week fact-tracking route checks whether the language work has become portable. The original aim remains to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it. Portability means that the learner can recognise the job in a fresh situation, retrieve a workable term, adapt the sentence and check the result. It does not require identical speed or wording every time. Begin with two library-event cards list audience, start time, duration, materials to bring and registration conditions, remove the completed model and note whether the learner asks for a word, a meaning distinction, a task reminder or no help at all.
Use a delayed retrieval check for a four-week fact-tracking route. Give the task cue rather than the target list: ask for the word that marks order, identifies a source, names a change, preserves a counterpoint or signals a recommendation, as appropriate. If evidence does not arrive, invite a plain-language explanation before offering a sound cue. Recovering the idea first keeps meaning in charge. After recall, the learner should use the word in a new sentence so that memory and application are checked together.
Change mode during a four-week fact-tracking route. Move from spoken explanation to a written line, from a card to a simple table, or from an individual response to a short partner discussion. Keep the underlying distinction whereas / however visible. A word that works only in the original format is not yet reliably transferable. If the switch creates too much load, restore one support—perhaps the evidence boxes—while leaving the vocabulary hidden. The aim is to isolate the new difficulty instead of treating every stumble as forgetting.
Fade support for a four-week fact-tracking route in an observable sequence: full worked example, incomplete frame, job cue, first-sound cue and independent attempt. Do not remove several supports together and then guess why performance fell. For two animal cards describe habitat, diet, body covering and one survival feature for an otter and a pangolin, record the lightest cue that produced an accurate response and whether the learner could explain the choice. Over time, the pattern should show which distinctions are secure, which need occasional retrieval and which still depend on adult language.
Families can support a four-week fact-tracking route with a short conversation, not a second classroom. Ask the child to explain one choice from a picture, notice, card or group decision. Teachers can preserve the same cue language across subjects. Learners can keep a tiny record of words that solved a real communication problem. None of these actions needs a daily score. The useful evidence is whether the child starts more readily, asks a more precise question, notices a mismatch or repairs a sentence with less help.
Use limits responsibly in a four-week fact-tracking route. Cambridge Primary Stage 3 is used as an official curriculum-stage label. This article does not convert it into a Singapore, British, American or other school level, and the cards are original teaching texts rather than Cambridge assessment materials. A child may show strong understanding through simpler words, or know a word but need time to retrieve it. Multilingual knowledge can support meaning even when English production is still developing. The exercises should therefore inform the next piece of teaching rather than serve as a universal maturity test. Follow the school’s language pathway and adapt examples so that the learner’s cultural knowledge is a resource, not an unspoken prerequisite.
The final progress check for a four-week fact-tracking route has four parts: accurate local meaning, clear reference, natural sentence fit and later independent use. Add a fifth only when relevant: the learner can explain why an alternative would be weaker. Keep the record descriptive—independent, small cue, full cue or not yet secure. Then choose one next experiment. Progress is the growing ability to use vocabulary for meaning, comprehension, writing and learning while carrying a sensible check of one’s own work.
CHAPTER 30 OF 30 · Transfer and check
30. Frequently asked questions and limits
Frequently asked questions and limits checks whether the language work has become portable. The original aim remains to hold the source of each fact steady, compare like with like and write a conclusion whose wording shows exactly which card supports it. Portability means that the learner can recognise the job in a fresh situation, retrieve a workable term, adapt the sentence and check the result. It does not require identical speed or wording every time. Begin with two plant cards describe leaf edge, stem, growing place and whether flowering was observed, remove the completed model and note whether the learner asks for a word, a meaning distinction, a task reminder or no help at all.
Use a delayed retrieval check for frequently asked questions and limits. Give the task cue rather than the target list: ask for the word that marks order, identifies a source, names a change, preserves a counterpoint or signals a recommendation, as appropriate. If according to does not arrive, invite a plain-language explanation before offering a sound cue. Recovering the idea first keeps meaning in charge. After recall, the learner should use the word in a new sentence so that memory and application are checked together.
Change mode during frequently asked questions and limits. Move from spoken explanation to a written line, from a card to a simple table, or from an individual response to a short partner discussion. Keep the underlying distinction interesting / relevant visible. A word that works only in the original format is not yet reliably transferable. If the switch creates too much load, restore one support—perhaps the evidence boxes—while leaving the vocabulary hidden. The aim is to isolate the new difficulty instead of treating every stumble as forgetting.
Fade support for frequently asked questions and limits in an observable sequence: full worked example, incomplete frame, job cue, first-sound cue and independent attempt. Do not remove several supports together and then guess why performance fell. For two transport cards give journey distance, usual duration, number of stops and accessibility information, record the lightest cue that produced an accurate response and whether the learner could explain the choice. Over time, the pattern should show which distinctions are secure, which need occasional retrieval and which still depend on adult language.
Families can support frequently asked questions and limits with a short conversation, not a second classroom. Ask the child to explain one choice from a picture, notice, card or group decision. Teachers can preserve the same cue language across subjects. Learners can keep a tiny record of words that solved a real communication problem. None of these actions needs a daily score. The useful evidence is whether the child starts more readily, asks a more precise question, notices a mismatch or repairs a sentence with less help.
Use limits responsibly in frequently asked questions and limits. Cambridge Primary Stage 3 is used as an official curriculum-stage label. This article does not convert it into a Singapore, British, American or other school level, and the cards are original teaching texts rather than Cambridge assessment materials. A child may show strong understanding through simpler words, or know a word but need time to retrieve it. Multilingual knowledge can support meaning even when English production is still developing. The exercises should therefore inform the next piece of teaching rather than serve as a universal maturity test. Follow the school’s language pathway and adapt examples so that the learner’s cultural knowledge is a resource, not an unspoken prerequisite.
The final progress check for frequently asked questions and limits has four parts: accurate local meaning, clear reference, natural sentence fit and later independent use. Add a fifth only when relevant: the learner can explain why an alternative would be weaker. Keep the record descriptive—independent, small cue, full cue or not yet secure. Then choose one next experiment. Progress is the growing ability to use vocabulary for meaning, comprehension, writing and learning while carrying a sensible check of one’s own work.
Sources, boundaries and review date
Sources reviewed 15 October 2026. Cambridge International pages establish the programme and subject names, flexibility, broad age range and Stage 1–6 structure. The evidence links inform vocabulary teaching principles; the original cases and answers are eduKate teaching material.
- Cambridge International: Cambridge Primary programme
- Cambridge International: Cambridge Primary curriculum
- Cambridge Primary English (0058)
- Cambridge Primary English endorsed resource list
- eduKateSG Vocabulary Learning Hub
- Education Endowment Foundation: Vocabulary in Action
- Goossens and colleagues: retrieval practice in primary-school vocabulary learning
- Goossens and colleagues: distributed and retrieval practice in primary-school vocabulary lessons
Evidence has limits. A result from one population, age group, task or classroom does not guarantee the same effect for every learner. Use retrieval and spacing as teaching principles to test, observe and adapt—not as promises of a particular score. Follow the school’s current curriculum, language pathway, safety guidance and assessment instructions.