Say ‘good at Mathematics’ when you mean that a child has ability in the subject: good normally combines with at before a subject, skill or activity. The actionable check is to ask whether the sentence evaluates ability; if it does, use good at, then place a noun, pronoun or -ing form after at.
In Punggol Primary 5 English tuition, this parent question connects adjective-preposition combinations, objects after prepositions, gerunds, comparisons, questions, editing and composition. ‘Good in’ is not banned from every possible sentence—‘She did good work in Mathematics class’ gives in its own context job—but it is not the standard direct ability pattern in ‘She is good at Mathematics’.
Parents searching for Primary 5 English tuition in Punggol, good at versus good in grammar help, English prepositions practice or an English tutor can use this focused guide. The MOE primary-school subjects and syllabuses page provides the official curriculum route, while the Punggol English Article Index remains the broad owner.
For another careful comparison of preposition frames, read Why Do We Say ‘Prefer Tea to Coffee’, Not ‘Prefer Tea Than Coffee’?. This guide stays with good at, the meanings of nearby good-in sentences and reliable transfer.
This guide keeps one parent question narrow so the established subject hub remains the broad owner. Use the five reading routes to begin at the exact misunderstanding, then move through worked examples, contrasts, diagnostics, useful practice and a proportionate parent decision.
For the broader Primary English route through grammar, comprehension, writing and editing, continue to the established subject index. Punggol English Article Index
Find your next learning step
ROUTE 1 · CHAPTERS 1–3
Answer and diagnose
Resolve the parent question and locate the first unstable idea.
ROUTE 2 · CHAPTERS 4–6
Build the mechanism
Connect words, representations or observations to the governing relationship.
ROUTE 3 · CHAPTERS 7–9
Test the boundary
Use near-misses and changed conditions so the rule remains accurate.
ROUTE 4 · CHAPTERS 10–12
Practise and explain
Work through varied examples, checks and school-style communication.
ROUTE 5 · CHAPTERS 13–15
Choose the next step
Use diagnostics, home practice, parent decisions and explicit FAQs.
Full chapter index · Start with the first checks · Existing English article index
Full chapter index
1–3 · Answer and diagnose
4–6 · Build the mechanism
7–9 · Test the boundary
10–12 · Practise and explain
13–15 · Choose the next step
1. The short answer: good at names the target of an ability
The target in this chapter is to use good at before a subject, skill or activity when describing ability. Begin with a prediction before offering a rule. Use this case: In ‘Mira is good at Mathematics’, Mathematics is the field in which her ability is being evaluated. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to use good at before a subject, skill or activity when describing ability. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. In ‘Mira is good at Mathematics’, Mathematics is the field in which her ability is being evaluated. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: ‘Mira is attentive in Mathematics lessons’ uses in to locate her behaviour within a setting, not to complete good. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is choosing in because Mathematics sounds like a school subject or field. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: sort twelve sentences into ability target, location and lesson context, then justify each preposition. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to ask whether the sentence evaluates ability or tells where something happens. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 1 with this transfer check: the learner completes an unseen good-at sentence and explains why in would change the construction. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
2. Why at works as a relationship marker
The target in this chapter is to understand at as part of the adjective-preposition combination good at rather than a literal place word. Begin with a prediction before offering a rule. Use this case: ‘Good at solving puzzles’ connects the judgement good to the activity solving puzzles. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to understand at as part of the adjective-preposition combination good at rather than a literal place word. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. ‘Good at solving puzzles’ connects the judgement good to the activity solving puzzles. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: ‘At the puzzle table’ uses at spatially; the same spelling can serve a different relationship. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is rejecting good at because at usually points to a place in early worksheets. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: underline the adjective-preposition unit and circle the noun phrase or -ing form that follows. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to teach the whole pattern good at instead of an isolated list of prepositions. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 2 with this transfer check: the learner identifies the target of ability in three new sentences. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
3. What can follow good at
The target in this chapter is to place nouns, pronouns and -ing forms after the preposition at. Begin with a prediction before offering a rule. Use this case: ‘Good at Science’, ‘good at it’ and ‘good at explaining results’ all place an object after at. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to place nouns, pronouns and -ing forms after the preposition at. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. ‘Good at Science’, ‘good at it’ and ‘good at explaining results’ all place an object after at. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: ‘Good to explain results’ is not the ordinary ability construction and needs a different sentence design. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is using the bare verb explain immediately after at. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: transform six verbs into -ing activities and substitute two pronouns without losing meaning. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to diagnose whether the problem is the preposition or the form that follows it. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 3 with this transfer check: the learner repairs ‘good at solve fractions’ and states both required changes. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
4. Good in can be possible when in has its own job
The target in this chapter is to avoid the false rule that the words good and in may never appear together. Begin with a prediction before offering a rule. Use this case: ‘The child did good work in Mathematics’ and ‘She feels good in this classroom’ give in a location or context role. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to avoid the false rule that the words good and in may never appear together. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. ‘The child did good work in Mathematics’ and ‘She feels good in this classroom’ give in a location or context role. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: For a direct judgement of ability, ‘She is good at Mathematics’ is the standard target construction. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is turning a helpful school rule into an absolute ban on good in. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: compare minimal pairs and paraphrase the relationship expressed by each preposition. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to prefer precise meaning over slogans such as good in is always wrong. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 4 with this transfer check: the learner accepts a grammatical good-in sentence while rejecting good in Mathematics as the target pattern. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
5. Good at and good with are not identical
The target in this chapter is to distinguish ability at a task from effective handling or relationships. Begin with a prediction before offering a rule. Use this case: ‘Good at mental calculation’ evaluates a task; ‘good with numbers’ suggests ease in handling numerical information. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to distinguish ability at a task from effective handling or relationships. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. ‘Good at mental calculation’ evaluates a task; ‘good with numbers’ suggests ease in handling numerical information. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: ‘Good with children’ commonly describes rapport, while ‘good at teaching children’ evaluates an activity. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is swapping at and with everywhere because both may sound complimentary. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: rewrite paired descriptions of tools, people, ideas and activities, noting the nuance. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to ask what the writer intends to praise rather than hunt for one keyword. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 5 with this transfer check: the learner explains the difference between good at coding and good with computers. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
6. Comparatives keep the same frame
The target in this chapter is to use better at and best at when comparing ability. Begin with a prediction before offering a rule. Use this case: ‘Ravi is better at fractions than at geometry’ keeps at before each compared target. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to use better at and best at when comparing ability. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. ‘Ravi is better at fractions than at geometry’ keeps at before each compared target. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: ‘Ravi performs better in quiet lessons’ gives in a setting function instead. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is changing to better in simply because the adjective changes form. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: build positive, comparative and superlative versions around the same activity. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to check parallel structure when two targets are compared. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 6 with this transfer check: the learner edits a comparison with mismatched prepositions. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
7. Questions and negatives
The target in this chapter is to preserve good at when auxiliaries alter word order. Begin with a prediction before offering a rule. Use this case: ‘What is Hana good at?’ and ‘Hana is not yet good at summarising’ retain the relationship. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to preserve good at when auxiliaries alter word order. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. ‘What is Hana good at?’ and ‘Hana is not yet good at summarising’ retain the relationship. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: ‘Where is Hana good?’ asks something different and may sound unnatural without context. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is dropping at when the target becomes a question word. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: turn statements into wh-questions, yes-no questions and gentle negative statements. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to separate word-order difficulty from preposition knowledge. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 7 with this transfer check: the learner asks and answers an ability question accurately. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
8. Avoid overgeneralising from weak in
The target in this chapter is to recognise that adjective-preposition combinations vary by adjective and register. Begin with a prediction before offering a rule. Use this case: School conversations may use ‘weak in algebra’, while ‘not confident with algebra’ or ‘needs help with algebra’ can be clearer. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to recognise that adjective-preposition combinations vary by adjective and register. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. School conversations may use ‘weak in algebra’, while ‘not confident with algebra’ or ‘needs help with algebra’ can be clearer. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: The pattern used with weak does not automatically determine the pattern used with good. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is assuming opposite adjectives must share the same preposition. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: collect adjective-preposition chunks and test them in complete sentences rather than analogy alone. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to consult reputable dictionaries when usage is uncertain and teach standard school English. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 8 with this transfer check: the learner rejects the claim that good in must be correct because weak in occurs. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
9. Use context to resolve ambiguous praise
The target in this chapter is to choose a construction that tells the reader exactly what ability is meant. Begin with a prediction before offering a rule. Use this case: ‘Aisha is good at presentations’ is clearer than ‘Aisha is good in class’ if speaking skill is the intended praise. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to choose a construction that tells the reader exactly what ability is meant. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. ‘Aisha is good at presentations’ is clearer than ‘Aisha is good in class’ if speaking skill is the intended praise. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: ‘Aisha is well behaved in class’ accurately locates behaviour but does not claim presentation ability. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is accepting a vague sentence merely because its grammar is possible. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: expand brief praise into evidence-based sentences with a target skill and one example. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to encourage specific feedback that a child can act on. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 9 with this transfer check: the learner revises vague praise without exaggerating performance. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
10. Composition character descriptions
The target in this chapter is to use good at naturally while showing rather than merely labelling ability. Begin with a prediction before offering a rule. Use this case: ‘Nabil was good at noticing small changes, so he spotted the loose screw before the shelf fell.’ links ability to action. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to use good at naturally while showing rather than merely labelling ability. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. ‘Nabil was good at noticing small changes, so he spotted the loose screw before the shelf fell.’ links ability to action. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: ‘Nabil was good in noticing’ is non-standard for the intended target relationship. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is repeating good at in every paragraph without evidence or variety. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: write one concise label, one action and one consequence for three characters. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to reward precise evidence instead of ornamental vocabulary. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 10 with this transfer check: the learner builds a believable character sentence around an ability. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
11. Editing for meaning, not sound alone
The target in this chapter is to test prepositions through the sentence’s intended relationship. Begin with a prediction before offering a rule. Use this case: In ‘Lena is good at planning but nervous in competitions’, both prepositions have distinct justified jobs. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to test prepositions through the sentence’s intended relationship. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. In ‘Lena is good at planning but nervous in competitions’, both prepositions have distinct justified jobs. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: Replacing both with the same preposition would flatten the contrast and may create an error. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is choosing whichever sequence sounds recently familiar. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: annotate adjective, preposition, object and intended relationship in an editing passage. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to ask the child to explain every correction rather than circle answers only. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 11 with this transfer check: the learner edits a mixed paragraph and preserves all intended meanings. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
12. Alternatives and register
The target in this chapter is to vary accurate descriptions without treating every alternative as interchangeable. Begin with a prediction before offering a rule. Use this case: ‘Skilled in data analysis’, ‘capable of careful analysis’ and ‘good at analysing data’ differ in tone and grammar. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to vary accurate descriptions without treating every alternative as interchangeable. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. ‘Skilled in data analysis’, ‘capable of careful analysis’ and ‘good at analysing data’ differ in tone and grammar. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: A formal noun phrase after skilled in does not license good in for the same meaning. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is copying the preposition from a synonym while changing the adjective. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: make a small phrase bank and use each frame in one suitable school context. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to teach combinations as units with examples, not universal substitution rules. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 12 with this transfer check: the learner selects an age-appropriate phrase and keeps its complement pattern. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
13. A seven-minute home routine
The target in this chapter is to retrieve the target pattern through short, spaced oral and written practice. Begin with a prediction before offering a rule. Use this case: Use family activities: good at baking, good with younger children, calm in the kitchen and skilled in decorating. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to retrieve the target pattern through short, spaced oral and written practice. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. Use family activities: good at baking, good with younger children, calm in the kitchen and skilled in decorating. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: On the next day, hide the phrase bank and ask for fresh examples with different nouns and -ing forms. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is copying twenty identical good-at sentences in one sitting. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: notice the meaning, choose the frame, explain it, write one pair and retest after two days. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to stop after accurate transfer and preserve the child’s own explanation. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 13 with this transfer check: the learner handles a delayed mixed set without a word bank. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
14. When Primary 5 English tuition has a clear job
The target in this chapter is to seek focused support when adjective-preposition choices remain unstable across grammar, editing and composition. Begin with a prediction before offering a rule. Use this case: Dated samples may show good in, interested on and afraid from across several tasks. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to seek focused support when adjective-preposition choices remain unstable across grammar, editing and composition. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. Dated samples may show good in, interested on and afraid from across several tasks. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: One self-corrected slip followed by stable transfer may need only a later check. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is buying a broad programme because one sentence sounded unusual. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: bring actual school sentences and ask how support will diagnose, contrast and retest the pattern. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to choose help that names the first weak link and measures independent transfer. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 14 with this transfer check: the parent can state the learning job and the evidence that would show improvement. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
15. Parent FAQs and final transfer
The target in this chapter is to consolidate ability targets, settings, nouns, -ing forms, comparisons and justified alternatives. Begin with a prediction before offering a rule. Use this case: The final task contrasts good at Mathematics, attentive in Mathematics lessons, good with numbers and skilled in calculation. Ask the learner to answer, justify the choice and point to the smallest piece of evidence that settles it. The first explanation is diagnostic evidence: it can reveal a vocabulary gap, an unsafe shortcut, a confused representation, a missing mechanism or difficulty communicating a sound idea clearly.
Build a relationship that predicts unfamiliar cases. For Primary 5 English, the learner should read the entire construction, identify the relationship carried by the preposition, preserve the intended meaning and test the sentence in a fresh context. The dependable idea here is to consolidate ability targets, settings, nouns, -ing forms, comparisons and justified alternatives. A remembered answer is only a starting point. Remove a familiar word, number or object and ask what remains true. Then ask what single change would genuinely require a different answer. Those questions turn recognition into control.
Work the central case in visible stages. The final task contrasts good at Mathematics, attentive in Mathematics lessons, good with numbers and skilled in calculation. First name what each word, digit, symbol, object or measurement represents. Next state the governing relationship in an ordinary sentence. Carry out one justified step at a time, then read the result back into the original question. A correct answer supported by an unsafe reason is not secure, because that reason may fail as soon as the surface details change.
Place a nearby case beside it: A possible good-in sentence must not erase the standard good-at ability pattern. Keep most features constant while changing the controlling condition, then preserve the relationship while changing the context. This double comparison prevents a recent keyword, visual pattern or memorised phrase from replacing thought. It also gives the learner accurate language for explaining where two routes agree and where they separate.
The tempting wrong route is memorising one correction without controlling the surrounding meanings. Treat that response as information rather than a character judgement. Ask what the learner noticed first, what rule or story they silently used and what evidence could make them revise it. Repair the earliest unsafe decision while preserving later work that was sound. Present a fresh near-miss immediately, so success cannot come from copying the model’s surface form.
Use this worked-practice sequence: answer the FAQs, edit a short profile and teach the relationship test aloud. Require three outputs each time: the answer, a reason and a check. Include a familiar item, a boundary case, a changed representation and a delayed cold item. Variation should be purposeful. The aim is to make the learner select the relationship independently, communicate it accurately and notice when a familiar-looking method is no longer allowed.
A useful parent move is to connect this narrow question to the established Punggol English hub. Praise a clear reason before speed. If the same weak link appears across several formats, keep two or three dated samples and describe the pattern precisely to the school teacher or tutor. A named pattern gives support a concrete job; broad labels such as weak in english hide the decision that actually needs repair.
Finish chapter 15 with this transfer check: the learner solves, explains and creates a boundary example independently. Remove the chapter heading and model, wait at least a day and change the setting. Ask the learner to solve, explain and invent one example that would make the answer different. If the reasoning remains stable, space the next review. If it collapses, return to the first unstable decision instead of adding a large pile of cloned questions.
Keep the emotional temperature low. A misconception that has become visible can now be improved. Let the learner compare the two routes aloud, revise one sentence or line of working, and say what cue will matter next time. End with one independent success. That small receipt is more informative than a long session that finishes with fatigue, and it gives the family a specific starting point for the next review.
Why is ‘good at Mathematics’ standard?
Good normally combines with at when the following phrase names the subject, skill or activity being evaluated. Mathematics is the target of ability in this sentence.
Is ‘good in’ always wrong?
No. In can have its own job in sentences such as ‘She did good work in Mathematics class’ or ‘She feels good in this room’. For the direct ability meaning, however, use ‘She is good at Mathematics’.
Can an -ing form follow good at?
Yes. A preposition is followed by an object, so ‘good at explaining’, ‘good at solving problems’ and ‘good at listening’ are standard constructions.
What is the difference between good at and good with?
Good at usually evaluates performance in a task or subject. Good with often suggests effective handling or rapport, as in ‘good with numbers’, ‘good with tools’ or ‘good with children’.
Does better use the same preposition?
Yes for the same ability relationship: ‘better at fractions’ and ‘best at editing’. Keep both sides parallel when comparing two targets.
What quick check can my child use?
Ask whether the phrase after the adjective names the target of ability. If it does, choose good at and check that the following noun, pronoun or -ing form is complete.
When can English tuition help?
Focused support is useful when adjective-preposition errors recur across grammar, editing and composition, especially when the child cannot explain the intended relationship. Bring dated examples so the first unstable pattern can be isolated.
