Primary 2 Mathematics is where mental calculation should begin to feel like reasoning rather than recitation. For families searching for Primary 2 Math tuition in Bukit Timah, Primary 2 Mathematics tuition in Singapore, P2 maths help, number bonds, times tables, multiplication and division, word problems, bar models or mental maths strategies, the most important question is not how many worksheets a child can finish. It is whether the child can see number structure well enough to choose an efficient route and explain why that route preserves the value of the calculation.
This guide keeps one central proposition from beginning to end: partitioning and compensation are not tricks for getting answers faster; they are early lessons in equivalence. A child who understands that 47 + 26 can become 47 + 20 + 6, or that 39 + 18 can become 40 + 18 – 1, is learning that a mathematical expression can be transformed without changing what it means. That is a small Primary 2 idea with a very long future.
The current Singapore Primary 2 Mathematics syllabus also asks for much more than fast addition. It includes numbers to 1000, addition and subtraction algorithms, mental calculation, multiplication and division within the 2, 3, 4, 5 and 10 tables, fractions, money, measurement, time, 2D and 3D shapes, and picture graphs with scales. This article therefore treats mental calculation as a spine: a way to build place value, flexibility, checking, explanation and strategy choice that later supports the rest of Primary Mathematics. Programme-specific Bukit Timah tuition information is kept separate; the specialist route is Bukit Timah Tutor Primary 2 Mathematics, while this page remains the deeper learning guide.
50-Second Router
| If this is what you are seeing | Start here |
|---|---|
| Your child still counts by ones for many sums | Read From counting to structure and Place value as the engine. |
| Your child can calculate but cannot explain the method | Read The invariant behind the strategy and How to teach explanation. |
| Your child rounds numbers but forgets to correct them | Read Compensation and Compensation failure modes. |
| Your child uses one method for every question | Read Strategy choice and the Mixed Strategy Lab. |
| Word problems cause blank starts | Read From words to relationships. |
| You want to know the official P2 scope | Read The current Singapore Primary 2 map. |
| You want a home routine without overloading the child | Read The six-week home plan. |
| You are deciding whether tuition is needed | Read When support is justified. |
| You want practice with answers and reasoning | Go to the Practice Laboratory. |
| You want the next stage | Use Primary 3 Mathematics and the Mathematics Article Directory. |
The Central Idea: Equivalent Routes, Same Quantity
A mental strategy is mathematically useful only if the child knows what has been preserved. Consider 47 + 26. Partitioning 26 into 20 and 6 changes the appearance of the calculation but not the quantity being added. The expression 47 + 26 and the sequence 47 + 20 + 6 represent the same total. Nothing has been invented and nothing has disappeared. The child has simply exposed the tens and ones structure inside 26.
Compensation makes the same principle more visible. In 39 + 18, changing 39 to 40 makes one part easier. But 40 is one larger than 39, so the transformed calculation is temporarily too large by one. Subtracting one restores equivalence: 40 + 18 – 1 = 57. A child who merely memorises “round then subtract one” may get several questions right, but a child who can say “I added one, so I must remove one later” has understood the invariant.
This is why mental calculation matters beyond speed. In later Mathematics, students constantly transform expressions while preserving meaning. They regroup terms, factorise, expand, rearrange equations, convert units, rewrite fractions, substitute equivalent forms and choose representations that make a problem easier. Primary 2 is not algebra, but the intellectual habit is already visible: change the representation, preserve the relationship.
Good mental calculation is not a race. It is the art of making a problem easier without changing the problem.
The Current Singapore Primary 2 Mathematics Map
The Ministry of Education Primary Mathematics syllabus, updated in October 2025, places Primary 2 inside a wider curriculum designed around concepts, skills, processes, metacognition and attitudes. The official Primary 2 content gives a useful boundary for parents: it tells us what belongs in the year and prevents tuition programmes from disguising premature acceleration as “advanced learning”.
Under Number and Algebra, Primary 2 includes numbers up to 1000; counting in tens and hundreds; place value in hundreds, tens and ones; reading and writing numerals and number words; comparing and ordering numbers; number patterns; odd and even numbers; addition and subtraction algorithms up to three digits; and mental calculation involving a three-digit number with ones, tens or hundreds. Multiplication and division introduce the 2, 3, 4, 5 and 10 tables, the division symbol, the relationship between multiplication and division, one-step word problems and mental calculation within those tables.
Fractions at this stage focus on fractions as parts of a whole, fraction notation and representation, comparing and ordering unit fractions and like fractions with denominators within the stated syllabus limit, and adding or subtracting like fractions within one whole. Money includes dollars and cents, decimal notation for money, comparison of amounts and conversion between dollars-and-cents notation and cents.
Measurement and Geometry include length in metres, mass in kilograms and grams, volume of liquid in litres, appropriate measurement units, comparison and ordering of quantities, time to the minute, duration in hours and minutes, conversions between hours-and-minutes and minutes, patterns with 2D shapes, and identification, description and classification of cube, cuboid, cone, cylinder and sphere. Statistics introduces reading and interpreting picture graphs with scales.
The official source is the MOE Primary Mathematics Syllabus. By the end of primary school, the SEAB PSLE Mathematics framework assesses recall and straightforward computation, interpretation and application across contexts, and mathematical reasoning including strategy selection. Primary 2 should therefore build a foundation for all three, not train only one narrow kind of speed.
| P2 strand | What the child should increasingly be able to do | Why mental structure matters |
|---|---|---|
| Whole numbers | Read, compare, decompose and operate with numbers up to 1000 | Partitioning depends on seeing hundreds, tens and ones as flexible units. |
| Addition & subtraction | Use algorithms and mental calculation appropriately | The child needs more than one route and must know when each route is efficient. |
| Multiplication & division | Connect equal groups, repeated structure and inverse relationships | Fluency is stronger when facts sit inside relationships instead of isolated chants. |
| Fractions | See parts of a whole and compare like structures | Partitioning a whole prepares the eye for fractional structure. |
| Money | Move accurately between dollars, cents and decimal notation | Place value and equivalence protect against decimal and unit confusion. |
| Measurement & time | Interpret quantities and convert within the taught units | Mental estimation helps the child notice impossible answers. |
| Geometry & data | Classify, describe and interpret representations | The same habit of reading structure before acting carries across strands. |
Primary 2 Is a Transition From Counting to Structure
Counting is not a bad strategy. It is an early strategy. A child who solves 8 + 5 by counting five steps forward has understood that addition increases a quantity. The problem appears when counting by ones remains the default long after the child has enough number knowledge to use larger structures. Counting one-by-one consumes attention. It is slow, easy to lose, and difficult to scale when numbers become larger or when a problem has several steps.
The goal is therefore not to ban counting. The goal is to give the child increasingly powerful units of thought. First the child may count objects. Then groups of objects. Then tens and ones. Later, the child may see 38 not as thirty-eight separate units but as 3 tens and 8 ones, or as 40 – 2, or as 20 + 18, or as 30 + 8. Each representation makes different calculations easier.
A flexible child can look at 58 + 27 and choose to add 20 then 7. Another child may make 60 by moving 2 from 27 to 58, leaving 25, so 60 + 25 = 85. Both are legitimate. The important development is not that every child must use the same “best” strategy. It is that the child begins to notice features of the numbers and select a strategy for a reason.
This is one place where adults accidentally weaken good mathematics. We sometimes reward the fastest visible answer and ignore the quality of the thinking that produced it. A child may answer quickly because a fact is well known, which is excellent. But for unfamiliar calculations, the stronger learner is often the one who pauses, notices structure, chooses a route, checks it, and can defend the choice. That is a more durable form of fluency.
Place Value Is the Engine
Partitioning works because the base-ten system is structured. When we write 347, the digits do not simply sit next to one another. The 3 represents 3 hundreds, the 4 represents 4 tens and the 7 represents 7 ones. A child who has only learned to read the number name may still have a fragile internal model. The real test is whether the child can decompose and recompose the number in several equivalent ways.
For example, 347 can be seen as 300 + 40 + 7. It can also be seen as 340 + 7, 300 + 47, 200 + 147, 350 – 3, or 400 – 53. Not all of these decompositions are equally useful for a Primary 2 learner, but the principle matters: a number is not trapped in one representation. Mental calculation becomes easier when the child can choose a decomposition that matches the operation.
Take 347 + 20. A child with strong place-value structure sees that only the tens quantity changes: 34 tens and 7 ones become 36 tens and 7 ones, producing 367. Take 347 + 200 and the hundreds component changes. Take 347 + 6 and the ones may cross a ten boundary. These are different cognitive jobs even though all are addition.
This suggests a useful diagnostic. Do not ask only, “Can you get the answer?” Ask, “What changed?” If the child says, “The 4 tens became 6 tens,” the mental model is visible. If the child repeatedly writes the whole algorithm for 347 + 20, the algorithm may be correct, but it may be hiding a weak sense of place value. A written algorithm is a powerful tool; it should not be the only tool.
A place-value conversation can be made concrete. Build 347 with base-ten materials or drawings: 3 hundred squares, 4 ten rods and 7 ones. Add two tens. Nothing happens to the ones or hundreds. Then write the same transformation symbolically. The physical and pictorial forms help the child connect the written digits to quantities rather than treating place value as a vocabulary exercise.
- Ask the child to show 426 in at least three ways.
- Ask which digit changes when 30 is added to 426 and why.
- Ask whether 426 + 30 is closer to 400, 450 or 500 before calculating.
- Ask the child to create a calculation where only the hundreds place changes.
- Ask the child to explain why adding 100 never changes the ones digit unless the original number itself is represented differently.
Number Bonds Are a Network, Not a List
Primary 1 number bonds remain important in Primary 2 because they become the micro-structures inside larger calculations. If 7 + 3 = 10 is instantly available, then 47 + 3, 97 + 3, 297 + 3 and 997 + 3 become easier to recognise. The child is reusing a relationship at a larger scale.
This is more powerful than memorising isolated facts because the child can derive what has not yet been memorised. If 8 + 2 = 10, then 8 + 5 can be thought of as 8 + 2 + 3 = 13. If 6 + 4 = 10, then 46 + 7 can become 46 + 4 + 3 = 53. A number bond acts like a bridge to a friendly benchmark.
Benchmarks such as 10, 20, 50 and 100 are especially useful because the base-ten system makes them easy to work with. A child should gradually develop a repertoire of complements: what goes with 1 to make 10, with 7 to make 10, with 38 to make 40, with 67 to make 70, and later with 86 to make 100. This is not a separate topic. It is the connective tissue of mental addition and subtraction.
The danger is turning number bonds into speed flashcards before the child understands the relationship. Fluency matters, but fluency built without structure can become brittle. A child may recall 7 + 3 but fail to use it inside 47 + 6. The stronger question is not merely “Do you know 7 + 3?” but “Where can 7 + 3 help you in this new calculation?”
Partitioning: Breaking a Number Without Breaking Its Value
Partitioning means decomposing a number into useful parts and operating on those parts while preserving the original value. For addition, one common route is to split the second addend into tens and ones. For subtraction, a child may subtract tens first, then ones. The strategy is transparent because every part can be accounted for.
Consider 47 + 26. Partition 26 into 20 and 6. Then 47 + 20 = 67 and 67 + 6 = 73. This route works because 20 + 6 is exactly 26. There is no approximation. The partition is simply a different representation of the same addend.
Now consider 63 – 28. Partition 28 into 20 and 8. First 63 – 20 = 43, then 43 – 8 = 35. That is mathematically sound. But a child who finds crossing the ten boundary difficult may use a more refined partition: subtract 3 to reach 60, then subtract the remaining 25, perhaps as 20 then 5. The quality of a partition depends on what the child can see and control.
Partitioning should therefore not become a rigid algorithm. “Always split the second number into tens and ones” is a useful beginning, not the final goal. Sometimes a different split is better. For 58 + 27, splitting 27 into 2 and 25 creates 60 + 25 = 85. That partition is chosen because 58 is close to 60. The strategy has become responsive rather than mechanical.
Alicia, Tricia and Kai Kai might all solve the same problem differently. Alicia may use tens then ones. Tricia may bridge to the next ten. Kai Kai may use compensation. The discussion becomes mathematically rich when they compare which steps were easiest to hold in mind, where an error could occur, and how each method can be checked. Different correct methods are not noise; they are evidence that the number relationships are becoming visible.
Partitioning addition: a progression
| Stage | Example | What the child is learning |
|---|---|---|
| Split tens and ones | 42 + 25 = 42 + 20 + 5 | A two-digit number can be decomposed without changing its value. |
| Bridge through a ten | 48 + 7 = 48 + 2 + 5 | A useful partition can be chosen to reach a friendly benchmark. |
| Recompose flexibly | 58 + 27 = 58 + 2 + 25 | Parts can be moved strategically while the total addition remains equivalent. |
| Scale the structure | 347 + 20 = 367 | Place value lets the child operate on units larger than one. |
| Explain equivalence | 47 + 26 and 47 + 20 + 6 are the same total | The method is justified, not merely performed. |
Partitioning subtraction: a progression
Subtraction deserves special attention because children can perform a sequence that looks similar to addition while losing track of direction. In 74 – 32, subtracting 30 then 2 is straightforward. In 74 – 38, subtracting 30 gives 44 and then subtracting 8 crosses a ten. If the child is not secure, a bridge strategy may be safer: 74 – 4 = 70, then subtract the remaining 34.
Another powerful subtraction interpretation is difference rather than take-away. The difference between 58 and 63 is 5. A child can “count up” from 58 to 60, then 60 to 63. This is especially efficient when the numbers are close. It also prepares the child to see subtraction as a relationship between two quantities, not only an instruction to remove objects.
The teaching goal is not to force one mental subtraction method. It is to expand the child’s strategy range while maintaining accuracy. The child should gradually know when a written algorithm is sensible, when a mental bridge is simpler, when counting up is efficient, and when a quick estimate can catch an impossible result.
Compensation: Make It Friendly, Then Repair the Change
Compensation starts from the observation that some numbers are almost friendly. Thirty-nine is almost forty. Forty-nine is almost fifty. Ninety-eight is almost one hundred. Instead of carrying an awkward number through the whole calculation, the child can temporarily replace it with a nearby benchmark and then compensate for the difference.
For 39 + 18, increase 39 by one to make 40. Now the calculation is 40 + 18 = 58. Because the transformed addend was one too large, subtract one: 58 – 1 = 57. The logic can be stated in one sentence: “I made 39 one bigger, so my answer became one too big; I remove one.”
For 52 – 19, it can be easier to subtract 20, giving 32, then add one back because 20 is one more than 19. The direction of the correction changes because the operation changed. This is exactly where superficial memorisation breaks. A child who merely remembers “round nineteen to twenty and then minus one” will make an error. The child must reason about whether the temporary calculation removed too much or too little.
Compensation can also preserve a difference by shifting both numbers equally. The difference between 63 and 28 is the same as the difference between 65 and 30, because both quantities increased by two. This “same difference” structure is elegant and later becomes useful in many forms of algebraic reasoning, but it should be introduced only when the child can explain why moving both quantities equally leaves the gap unchanged.
The deepest lesson is again invariance. In addition, one quantity may be adjusted and then the total corrected. In subtraction-as-difference, both quantities may be shifted equally so the gap remains fixed. The child is learning to manipulate representation while tracking what must stay unchanged.
Compensation failure modes
| What the child does | Likely issue | Repair question |
|---|---|---|
| Rounds 39 to 40 in 39 + 18 and answers 58 | Forgets that the transformed problem is one larger | What did you change? Did that make the total too big or too small? |
| Does 52 – 19 as 52 – 20 – 1 | Correction direction is reversed | If you removed 20 instead of 19, did you remove too much or too little? |
| Compensates even when partitioning is simpler | Strategy has become a rule instead of a choice | Which route has fewer mental steps for these particular numbers? |
| Cannot explain why both numbers may shift in a difference problem | Procedure has outrun understanding | If both points move two steps right on a number line, what happens to the gap? |
| Gives a correct answer but cannot reconstruct the steps | Working memory may be overloaded or the method may be guessed | Show the change on a number line or with place-value blocks. |
Strategy Choice Is the Real Fluency
Fluency is often misunderstood as speed alone. A more complete view includes accuracy, efficiency, flexibility and appropriate strategy selection. The child who knows three methods but always uses the longest one is not yet flexible. The child who jumps between methods randomly is not yet controlled. The child who chooses a route because of the number structure is developing fluency.
Consider 48 + 7. Partitioning 7 into 2 and 5 to make 50 is probably more efficient than splitting 7 into 5 and 2 without reference to a benchmark. Consider 42 + 25. Adding 20 then 5 is direct. Consider 99 + 36. Compensation through 100 is attractive. Consider 347 + 30. Place-value reasoning is simplest: add three tens. The “best” method is local to the numbers.
A useful teaching routine is to ask three questions after the answer: Why this method? What other method could work? Which method would you choose next time? These questions transform a worksheet from a sequence of answers into a sequence of decisions.
This is also why excessive timed drilling can be counterproductive when used as the only mode of practice. Timed practice can help consolidate facts that are already understood, but if every task rewards the first answer produced, children may cling to familiar counting procedures rather than experiment with more efficient structure. Strategy development needs moments of deliberate comparison.
Mental Calculation and Written Algorithms Should Support Each Other
Mental calculation is not a campaign against written working. Written algorithms are essential tools. The educational question is whether the child knows when a written algorithm is necessary and whether the written steps still carry meaning. A child who can do 347 + 20 mentally should not be forced to write a full vertical algorithm merely to prove seriousness. A child facing a more complex multi-step calculation should not be pressured to hold everything in the head.
The ideal relationship is complementary. Mental methods develop number sense, estimation and flexible decomposition. Written methods provide reliability, auditability and lower working-memory demand. Estimation checks the scale of an answer. A strong learner can move among these modes rather than confusing one mode with mathematical ability itself.
For example, before calculating 286 + 197 with a written algorithm, a child might estimate that the answer is a little under 500. After obtaining 483, the estimate supports plausibility. If the algorithm produced 383 or 583, the child has a reason to investigate. The mental estimate becomes a safety system around the written procedure.
Multiplication and Division: Relationships Before Chanting
Primary 2 introduces multiplication and division within the 2, 3, 4, 5 and 10 tables. Automatic recall is valuable because later problems become unnecessarily difficult when a child must reconstruct every basic fact. But automaticity is strongest when it grows from meaning.
If 4 groups of 3 make 12, the child should be able to connect 4 × 3 = 12 with 12 ÷ 3 = 4 and 12 ÷ 4 = 3. These are not three unrelated facts. They describe one multiplicative structure from different directions. Equal groups, arrays, skip patterns and number-line jumps can all support that structure.
Partitioning also appears in multiplication. A child who knows 5 × 6 may derive 4 × 6 by removing one group of 6, or derive 6 × 6 by adding one group. This should not replace learning the required facts, but it gives the facts internal connections. When memory fails, the network can rebuild the answer.
The same principle protects division from becoming “reverse times tables” without meaning. Twelve divided by three can mean “How many groups of three fit in twelve?” or “If twelve is shared equally into three groups, how many are in each group?” The numerical result may match, but the situation model differs. P2 word problems are where this distinction begins to matter.
Word Problems: Read the Relationship Before Choosing the Operation
A child can be fast at arithmetic and still struggle with word problems because a word problem is not merely a calculation written in English. It requires the learner to identify quantities, relationships, what is known, what is unknown and what operation represents the situation. The arithmetic comes after the model.
For Primary 2, the first discipline is to delay computation long enough to understand the situation. “Alicia has 38 stickers. Tricia gives her 17 more. How many stickers does Alicia have now?” is an increase structure. “Alicia has 38 stickers. Tricia has 17 stickers. How many more does Alicia have?” is a comparison structure. The same numbers appear, but the relationship and operation differ.
This is where drawings and bar models can help. A bar model is useful when it makes the relationship visible; it is not a decorative requirement that must be drawn for every easy question. For a child who loses track of comparison language, a longer bar and shorter bar can stabilise the meaning of “how many more”. For a simple known fact, drawing a full model may add unnecessary load. Representation should serve thinking.
A good P2 problem-solving routine can be compact: identify the quantities, state the relationship in words, choose or draw a representation if needed, write the operation, calculate, then read the answer back into the story. The final step matters. “73” is not a complete interpretation; “Alicia has 73 stickers” reconnects the number to the situation.
Parents can diagnose word-problem failure by separating four possible causes: language, representation, operation choice and arithmetic. If the child misreads “fewer than”, extra arithmetic drills may not help. If the child chooses the correct operation but makes a regrouping error, the representation may be fine. If the child gets the calculation right but answers the wrong question, attention to the requested quantity is the problem. The intervention should match the failure.
| Failure layer | What it looks like | Useful response |
|---|---|---|
| Language | Child cannot explain what a phrase means | Paraphrase the sentence and act out or draw the quantities. |
| Representation | Child understands the words but cannot organise the quantities | Use a simple bar, part-whole drawing, table or labelled sketch. |
| Operation choice | Child has the quantities but chooses the wrong relationship | Ask what is changing: combining, separating, comparing, grouping or sharing. |
| Arithmetic | Model and operation are correct, computation fails | Repair the specific number relationship or algorithm. |
| Answer interpretation | Child calculates correctly but gives an irrelevant response | Return to the question and state the answer with units/context. |
Mental Calculation Across Money, Measurement and Time
Mental structure should not remain trapped inside “number” worksheets. It becomes more useful when transferred into the rest of the P2 syllabus. Money is an obvious bridge because Singapore currency naturally uses place-value and complement ideas. If an item costs $2.70 and a child pays $3.00, the difference is 30 cents. The child can count up from $2.70 to $3.00 rather than forcing a written subtraction algorithm.
Measurement provides estimation opportunities. If one object is 1 metre long and another is 30 centimetres, a child should know that simply adding the numerals 1 and 30 is meaningless unless the units are reconciled. At P2, the official syllabus does not require every possible conversion across all units, so adults should stay within taught expectations, but the broader habit is worth cultivating: numbers describe quantities, and units are part of the quantity.
Time also rewards structure. If an activity starts at 2:35 pm and lasts 20 minutes, the child can partition the duration: 25 minutes would reach 3:00 pm, so 20 minutes reaches 2:55 pm. If the duration is 35 minutes, the child can bridge to the next hour and continue. This is the same mental architecture appearing in a different context.
Picture graphs with scales create another transfer. When one picture represents two objects, children must treat the icon as a unit larger than one. This is conceptually related to grouping in multiplication. Good Mathematics instruction helps the child notice these recurring structures rather than treating every chapter as an isolated island.
Fractions: Partitioning a Whole Changes the Meaning of the Parts
The word “partitioning” appears in another important sense when children learn fractions: dividing a whole into equal parts. This is not the same procedure as decomposing 26 into 20 + 6, but the shared idea is that a quantity can be represented through parts. The condition of equal parts is essential for fractions.
A child who understands that one whole can be partitioned into four equal parts can interpret one part as one quarter. Two quarters then describe two of those equal parts. Comparing like fractions becomes easier because the denominator tells us the size structure of the parts and the numerator tells us how many of those parts are considered.
Mental calculation habits help here because the child is already learning to ask what has been preserved. Cutting a pizza into more equal pieces changes the size of each piece but not the amount of pizza in the whole. Later, equivalent fractions formalise this kind of invariance more deeply. At P2, the focus should remain concrete and pictorial enough for the child to see the parts, not merely manipulate symbols.
Concrete → Pictorial → Abstract Is a Translation Chain
Young learners often appear to understand a concept in one representation and then fail when the representation changes. A child may successfully add base-ten blocks but become confused by the same problem written only in numerals. That is not necessarily carelessness. It may indicate that the translation between representations is incomplete.
A robust sequence moves among concrete objects, pictorial representations and abstract symbols. Concrete does not mean childish; it means the quantity can be manipulated physically. Pictorial forms include drawings, ten frames, number lines, part-whole diagrams and bar models. Abstract forms include numerals and equations. The educational power comes from making the links explicit.
For 38 + 7, build 38 as three tens and eight ones. Add seven ones. Regroup ten ones into one ten. Then draw the change. Then write 38 + 2 + 5 = 45. The concrete regrouping, pictorial bridge and symbolic partition all describe the same quantity transformation. When these representations agree, the child is less likely to treat symbolic steps as arbitrary rules.
Research summaries on mental computation repeatedly emphasise strategy development, discussion and justification rather than a single mandated procedure. The ERIC record for Ann Heirdsfield’s Teaching Mental Computation Strategies in Early Mathematics highlights the value of children exploring, discussing and justifying strategies as a way to develop number sense. The classroom implication is simple: a mental method should be explainable.
The Open Number Line as a Thinking Tool
An open number line is useful because it records mental jumps without prescribing every mark. For 47 + 26, a child may start at 47, jump +20 to 67, then +6 to 73. Another may jump +3 to 50, then +23 to 73. The line makes the route visible while preserving the child’s strategy.
For subtraction, the number line can support both take-away and difference. To find 63 – 58, start at 58 and jump to 60, then to 63. The total upward distance is five. This often feels more natural than taking 58 away from 63 one part at a time. The visual gap makes the “difference” meaning explicit.
The number line should eventually become optional. If the child needs it, use it. If the child can hold the jumps mentally and explain them accurately, forcing a drawing for every question adds work without adding understanding. Scaffolds are successful when they can be removed.
How to Teach Explanation Without Turning Mathematics Into an Oral Exam
Asking children to explain thinking is valuable, but over-questioning can make a simple calculation exhausting. The goal is not to demand a speech after every answer. The goal is to sample the child’s reasoning often enough to know whether the method has meaning.
A compact explanation protocol works well: “What did you change?” “Why was that allowed?” “How do you know the answer is reasonable?” For 39 + 18, the child might say, “I changed 39 to 40 because 40 is easier. That added one extra, so I subtracted one at the end. Fifty-seven is reasonable because 40 + 20 would be 60.” Three sentences reveal transformation, compensation and estimation.
Different children need different levels of language support. Alicia may explain fluently. Tricia may understand but need sentence starters. Kai Kai may first point to a number line and use short phrases. The teacher should not confuse verbal fluency with mathematical understanding, but should gradually help every child connect actions, diagrams and equations to mathematical language.
- “I split ___ into ___ and ___ because …”
- “I changed ___ to ___ because …”
- “That made the answer ___ too big / too small, so …”
- “I know this is reasonable because …”
- “Another method would be …”
- “This method is easier here because …”
Error Taxonomy: Wrong Answers Have Different Causes
If every wrong answer is treated as “carelessness”, teaching loses diagnostic resolution. Primary 2 errors can arise from very different sources. A child may not know a fact, may know the fact but misread place value, may understand the quantities but lose a step in working memory, may select an unsuitable strategy, may reverse a compensation correction, or may misunderstand the language of a word problem.
The practical consequence is that correction should begin with classification. If a child writes 347 + 20 = 349, the issue is not “addition” in general; the child may be treating 20 as two ones. If a child writes 39 + 18 = 58 after compensation, the issue is not place value; the correction step is missing. If a child solves 52 – 19 as 52 – 20 – 1, the child may understand the rounding but not the direction of compensation.
A small diagnostic conversation can therefore outperform a page of repeated questions. Ask the child to solve one problem aloud, draw the route, solve a changed version and then return to the original later. The teacher is looking for the earliest point where meaning breaks. Repair that point, then rebuild the chain.
| Error family | Example | Likely root | First repair |
|---|---|---|---|
| Place-value confusion | 347 + 20 = 349 | 20 treated as 2 ones | Rebuild hundreds/tens/ones physically and pictorially. |
| Lost partition | 47 + 26 → 47 + 20 = 67 and stops | One part disappears from working memory | Record the partition before calculating. |
| Compensation omitted | 39 + 18 → 58 | Temporary change not repaired | Ask how the transformed problem differs from the original. |
| Compensation reversed | 52 – 19 → 31 | Removed 20, then removed one more | Use a number line to show that subtracting 20 removed too much. |
| Counting dependence | 46 + 7 counted one by one | Number bonds not available for bridging | Practise complements to the next ten in context. |
| Strategy rigidity | Uses vertical algorithm for 347 + 20 | No strategy selection habit | Compare mental and written routes; ask which has fewer steps. |
| Word-problem language | Adds when the question asks how many fewer | Relationship not understood | Represent the two quantities before choosing an operation. |
| Unit loss | Writes 2 instead of 2 kg | Number detached from quantity | Require answer-with-unit and estimate plausibility. |
Three Students, Three Strategies: What a Small Group Can Reveal
A three-student learning setting can be educationally useful when it is used as an observation laboratory rather than simply a smaller lecture. Three children solving one problem may expose three different strategies, three different misconceptions or three different levels of confidence. The teacher can then compare reasoning without losing sight of individual work.
Imagine 58 + 27. Alicia adds 20, then 7. Tricia moves 2 from 27 to 58 and calculates 60 + 25. Kai Kai writes the vertical algorithm. All three are correct. The teacher’s task is not to declare a winner. It is to ask what each method makes easy, where each might fail, and which method each child would choose for 58 + 42, 99 + 27 or 347 + 20.
The group becomes powerful when students first think independently, then compare. If the strongest child announces the method too early, the others may imitate rather than reason. If the teacher rescues every hesitation, students learn that waiting produces hints. Productive small-group teaching therefore includes silence, individual attempt time, targeted questions and deliberate withdrawal of support.
The same logic explains why class size alone does not guarantee quality. A three-student class can still be poor if every child copies the same worksheet silently and receives answer-level correction. The value comes from visibility: the teacher can see the route, identify the earliest weak link, intervene briefly, and then return responsibility to the child.
When Primary 2 Tuition Is Actually Justified
Tuition is not automatically necessary in Primary 2. Many children learn effectively through school, home reading of feedback and modest independent practice. A weekly programme is justified only when there is a clear learning job that the current environment is not resolving efficiently.
Persistent counting dependence, repeated place-value errors, weak multiplication/division meaning, recurring word-problem misinterpretation or a pattern of corrections that do not transfer to changed questions may justify additional support. One weak worksheet does not. One unfamiliar chapter does not. A parent hearing that “everyone in Bukit Timah has tuition” does not.
The best decision is evidence-based. Look across classwork, homework, school feedback and live attempts. Ask whether the same error recurs. Check whether a short targeted repair changes future performance. If the child can learn, retrieve and transfer the idea through school alone, adding tuition may create workload without adding much value.
For families who want a specialist commercial route, keep that decision separate from this learning article. The bounded Bukit Timah Mathematics programme information belongs at Bukit Timah Tutor Primary 2 Mathematics. The role of this eduKateSingapore page is to help a reader understand the mathematics well enough to make a better decision.
A Six-Week Home Plan: Small, Specific, Sustainable
A home plan should not attempt to recreate school. Ten to fifteen focused minutes can be more useful than an hour of tired repetition if the activity has one clear purpose. The sequence below is designed to build from relationships to strategy choice. It is not a substitute for the child’s school programme and should be adjusted to actual school progress.
| Week | Focus | Example activity | What success looks like |
|---|---|---|---|
| 1 | Place value and complements | Build numbers to 1000; ask what changes when adding 10, 20, 100; practise complements to the next ten | Child identifies tens/hundreds changes without counting every unit. |
| 2 | Partitioning addition | Solve 2-digit additions by tens-and-ones and by bridging | Child can account for every part and explain why the total is preserved. |
| 3 | Partitioning subtraction | Use take-away and counting-up routes | Child can choose a stable route and avoids losing a partition. |
| 4 | Compensation | Work with numbers near 10, 20, 50, 100 | Child states whether the temporary answer is too large or too small and corrects in the right direction. |
| 5 | Strategy choice | Mix questions without naming the method | Child chooses a method based on number structure rather than chapter cues. |
| 6 | Transfer | Use money, time and simple word problems | Child carries mental structure into context and checks units/meaning. |
Do not increase difficulty merely because the child succeeds once. First remove support, then delay the question, then change the numbers, then change the context. A strategy is becoming stable when it survives those changes.
A useful weekly review is “three old, two recent, one new”. The three old questions test retrieval after delay. The two recent questions consolidate current learning. The one new question creates a small amount of productive novelty. This prevents the common illusion created by doing twenty nearly identical questions in one sitting.
Practice Laboratory: Partition, Compensate, Choose, Explain
The practice below is deliberately organised by reasoning job, not by difficulty label. Answers are included with route notes so parents and teachers can diagnose the method, not just the final number. A child does not need to complete everything. Select a small set, remove the route hint once the structure is stable, and revisit changed versions after a delay.
Lab A — Partitioning addition
A1. 34 + 25 = 59. A direct partition is 34 + 20 + 5. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 20 + 5 is exactly 25, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A2. 42 + 36 = 78. A direct partition is 42 + 30 + 6. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 30 + 6 is exactly 36, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A3. 53 + 24 = 77. A direct partition is 53 + 20 + 4. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 20 + 4 is exactly 24, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A4. 61 + 28 = 89. A direct partition is 61 + 20 + 8. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 20 + 8 is exactly 28, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A5. 27 + 42 = 69. A direct partition is 27 + 40 + 2. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 40 + 2 is exactly 42, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A6. 46 + 33 = 79. A direct partition is 46 + 30 + 3. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 30 + 3 is exactly 33, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A7. 55 + 27 = 82. A direct partition is 55 + 20 + 7. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 20 + 7 is exactly 27, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A8. 68 + 21 = 89. A direct partition is 68 + 20 + 1. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 20 + 1 is exactly 21, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A9. 72 + 16 = 88. A direct partition is 72 + 10 + 6. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 10 + 6 is exactly 16, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A10. 43 + 29 = 72. A direct partition is 43 + 20 + 9. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 20 + 9 is exactly 29, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A11. 47 + 26 = 73. A direct partition is 47 + 20 + 6. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 20 + 6 is exactly 26, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A12. 38 + 45 = 83. A direct partition is 38 + 40 + 5. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 40 + 5 is exactly 45, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A13. 56 + 18 = 74. A direct partition is 56 + 10 + 8. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 10 + 8 is exactly 18, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A14. 64 + 27 = 91. A direct partition is 64 + 20 + 7. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 20 + 7 is exactly 27, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A15. 29 + 54 = 83. A direct partition is 29 + 50 + 4. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 50 + 4 is exactly 54, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A16. 35 + 48 = 83. A direct partition is 35 + 40 + 8. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 40 + 8 is exactly 48, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A17. 57 + 34 = 91. A direct partition is 57 + 30 + 4. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 30 + 4 is exactly 34, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A18. 62 + 29 = 91. A direct partition is 62 + 20 + 9. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 20 + 9 is exactly 29, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A19. 48 + 37 = 85. A direct partition is 48 + 30 + 7. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 30 + 7 is exactly 37, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
A20. 73 + 18 = 91. A direct partition is 73 + 10 + 8. Add the tens first, then the ones. The teaching point is not that this route is compulsory; it is that 10 + 8 is exactly 18, so the transformed calculation preserves the original total. Ask the child whether a bridge-to-ten route would be shorter and why.
Lab B — Bridge to a friendly ten
B1. 48 + 7 = 55. 48 needs 2 to reach 50. Partition 7 into 2 and 5; then calculate 48 + 2 + 5. This is useful because the first step lands on a multiple of ten. The child should be able to say that 2 + 5 still equals 7, so no quantity has been lost or created.
B2. 57 + 8 = 65. 57 needs 3 to reach 60. Partition 8 into 3 and 5; then calculate 57 + 3 + 5. This is useful because the first step lands on a multiple of ten. The child should be able to say that 3 + 5 still equals 8, so no quantity has been lost or created.
B3. 36 + 9 = 45. 36 needs 4 to reach 40. Partition 9 into 4 and 5; then calculate 36 + 4 + 5. This is useful because the first step lands on a multiple of ten. The child should be able to say that 4 + 5 still equals 9, so no quantity has been lost or created.
B4. 69 + 6 = 75. 69 needs 1 to reach 70. Partition 6 into 1 and 5; then calculate 69 + 1 + 5. This is useful because the first step lands on a multiple of ten. The child should be able to say that 1 + 5 still equals 6, so no quantity has been lost or created.
B5. 78 + 5 = 83. 78 needs 2 to reach 80. Partition 5 into 2 and 3; then calculate 78 + 2 + 3. This is useful because the first step lands on a multiple of ten. The child should be able to say that 2 + 3 still equals 5, so no quantity has been lost or created.
B6. 47 + 8 = 55. 47 needs 3 to reach 50. Partition 8 into 3 and 5; then calculate 47 + 3 + 5. This is useful because the first step lands on a multiple of ten. The child should be able to say that 3 + 5 still equals 8, so no quantity has been lost or created.
B7. 28 + 7 = 35. 28 needs 2 to reach 30. Partition 7 into 2 and 5; then calculate 28 + 2 + 5. This is useful because the first step lands on a multiple of ten. The child should be able to say that 2 + 5 still equals 7, so no quantity has been lost or created.
B8. 59 + 4 = 63. 59 needs 1 to reach 60. Partition 4 into 1 and 3; then calculate 59 + 1 + 3. This is useful because the first step lands on a multiple of ten. The child should be able to say that 1 + 3 still equals 4, so no quantity has been lost or created.
B9. 67 + 6 = 73. 67 needs 3 to reach 70. Partition 6 into 3 and 3; then calculate 67 + 3 + 3. This is useful because the first step lands on a multiple of ten. The child should be able to say that 3 + 3 still equals 6, so no quantity has been lost or created.
B10. 39 + 8 = 47. 39 needs 1 to reach 40. Partition 8 into 1 and 7; then calculate 39 + 1 + 7. This is useful because the first step lands on a multiple of ten. The child should be able to say that 1 + 7 still equals 8, so no quantity has been lost or created.
B11. 58 + 27 = 85. 58 needs 2 to reach 60. Partition 27 into 2 and 25; then calculate 58 + 2 + 25. This is useful because the first step lands on a multiple of ten. The child should be able to say that 2 + 25 still equals 27, so no quantity has been lost or created.
B12. 49 + 36 = 85. 49 needs 1 to reach 50. Partition 36 into 1 and 35; then calculate 49 + 1 + 35. This is useful because the first step lands on a multiple of ten. The child should be able to say that 1 + 35 still equals 36, so no quantity has been lost or created.
B13. 68 + 25 = 93. 68 needs 2 to reach 70. Partition 25 into 2 and 23; then calculate 68 + 2 + 23. This is useful because the first step lands on a multiple of ten. The child should be able to say that 2 + 23 still equals 25, so no quantity has been lost or created.
B14. 79 + 14 = 93. 79 needs 1 to reach 80. Partition 14 into 1 and 13; then calculate 79 + 1 + 13. This is useful because the first step lands on a multiple of ten. The child should be able to say that 1 + 13 still equals 14, so no quantity has been lost or created.
B15. 38 + 27 = 65. 38 needs 2 to reach 40. Partition 27 into 2 and 25; then calculate 38 + 2 + 25. This is useful because the first step lands on a multiple of ten. The child should be able to say that 2 + 25 still equals 27, so no quantity has been lost or created.
B16. 57 + 26 = 83. 57 needs 3 to reach 60. Partition 26 into 3 and 23; then calculate 57 + 3 + 23. This is useful because the first step lands on a multiple of ten. The child should be able to say that 3 + 23 still equals 26, so no quantity has been lost or created.
B17. 46 + 38 = 84. 46 needs 4 to reach 50. Partition 38 into 4 and 34; then calculate 46 + 4 + 34. This is useful because the first step lands on a multiple of ten. The child should be able to say that 4 + 34 still equals 38, so no quantity has been lost or created.
B18. 29 + 45 = 74. 29 needs 1 to reach 30. Partition 45 into 1 and 44; then calculate 29 + 1 + 44. This is useful because the first step lands on a multiple of ten. The child should be able to say that 1 + 44 still equals 45, so no quantity has been lost or created.
B19. 69 + 24 = 93. 69 needs 1 to reach 70. Partition 24 into 1 and 23; then calculate 69 + 1 + 23. This is useful because the first step lands on a multiple of ten. The child should be able to say that 1 + 23 still equals 24, so no quantity has been lost or created.
B20. 88 + 17 = 105. 88 needs 2 to reach 90. Partition 17 into 2 and 15; then calculate 88 + 2 + 15. This is useful because the first step lands on a multiple of ten. The child should be able to say that 2 + 15 still equals 17, so no quantity has been lost or created.
Lab C — Compensation in addition
C1. 39 + 18 = 57. Increase 39 by 1 to make 40. The easier temporary calculation is 40 + 18 = 58. Because the first addend was made 1 too large, subtract 1 from the temporary total. Ask the child to estimate first: the answer should be close to 58, but slightly smaller.
C2. 49 + 27 = 76. Increase 49 by 1 to make 50. The easier temporary calculation is 50 + 27 = 77. Because the first addend was made 1 too large, subtract 1 from the temporary total. Ask the child to estimate first: the answer should be close to 77, but slightly smaller.
C3. 59 + 34 = 93. Increase 59 by 1 to make 60. The easier temporary calculation is 60 + 34 = 94. Because the first addend was made 1 too large, subtract 1 from the temporary total. Ask the child to estimate first: the answer should be close to 94, but slightly smaller.
C4. 29 + 46 = 75. Increase 29 by 1 to make 30. The easier temporary calculation is 30 + 46 = 76. Because the first addend was made 1 too large, subtract 1 from the temporary total. Ask the child to estimate first: the answer should be close to 76, but slightly smaller.
C5. 69 + 25 = 94. Increase 69 by 1 to make 70. The easier temporary calculation is 70 + 25 = 95. Because the first addend was made 1 too large, subtract 1 from the temporary total. Ask the child to estimate first: the answer should be close to 95, but slightly smaller.
C6. 79 + 16 = 95. Increase 79 by 1 to make 80. The easier temporary calculation is 80 + 16 = 96. Because the first addend was made 1 too large, subtract 1 from the temporary total. Ask the child to estimate first: the answer should be close to 96, but slightly smaller.
C7. 89 + 24 = 113. Increase 89 by 1 to make 90. The easier temporary calculation is 90 + 24 = 114. Because the first addend was made 1 too large, subtract 1 from the temporary total. Ask the child to estimate first: the answer should be close to 114, but slightly smaller.
C8. 99 + 17 = 116. Increase 99 by 1 to make 100. The easier temporary calculation is 100 + 17 = 117. Because the first addend was made 1 too large, subtract 1 from the temporary total. Ask the child to estimate first: the answer should be close to 117, but slightly smaller.
C9. 38 + 26 = 64. Increase 38 by 2 to make 40. The easier temporary calculation is 40 + 26 = 66. Because the first addend was made 2 too large, subtract 2 from the temporary total. Ask the child to estimate first: the answer should be close to 66, but slightly smaller.
C10. 48 + 35 = 83. Increase 48 by 2 to make 50. The easier temporary calculation is 50 + 35 = 85. Because the first addend was made 2 too large, subtract 2 from the temporary total. Ask the child to estimate first: the answer should be close to 85, but slightly smaller.
C11. 58 + 27 = 85. Increase 58 by 2 to make 60. The easier temporary calculation is 60 + 27 = 87. Because the first addend was made 2 too large, subtract 2 from the temporary total. Ask the child to estimate first: the answer should be close to 87, but slightly smaller.
C12. 68 + 24 = 92. Increase 68 by 2 to make 70. The easier temporary calculation is 70 + 24 = 94. Because the first addend was made 2 too large, subtract 2 from the temporary total. Ask the child to estimate first: the answer should be close to 94, but slightly smaller.
C13. 78 + 19 = 97. Increase 78 by 2 to make 80. The easier temporary calculation is 80 + 19 = 99. Because the first addend was made 2 too large, subtract 2 from the temporary total. Ask the child to estimate first: the answer should be close to 99, but slightly smaller.
C14. 88 + 15 = 103. Increase 88 by 2 to make 90. The easier temporary calculation is 90 + 15 = 105. Because the first addend was made 2 too large, subtract 2 from the temporary total. Ask the child to estimate first: the answer should be close to 105, but slightly smaller.
C15. 98 + 23 = 121. Increase 98 by 2 to make 100. The easier temporary calculation is 100 + 23 = 123. Because the first addend was made 2 too large, subtract 2 from the temporary total. Ask the child to estimate first: the answer should be close to 123, but slightly smaller.
Lab D — Partitioning subtraction
D1. 74 – 32 = 42. One route is 74 – 30 = 44, then 44 – 2 = 42. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 32 and preserves the subtraction direction.
D2. 86 – 24 = 62. One route is 86 – 20 = 66, then 66 – 4 = 62. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 24 and preserves the subtraction direction.
D3. 95 – 43 = 52. One route is 95 – 40 = 55, then 55 – 3 = 52. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 43 and preserves the subtraction direction.
D4. 67 – 25 = 42. One route is 67 – 20 = 47, then 47 – 5 = 42. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 25 and preserves the subtraction direction.
D5. 82 – 31 = 51. One route is 82 – 30 = 52, then 52 – 1 = 51. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 31 and preserves the subtraction direction.
D6. 73 – 26 = 47. One route is 73 – 20 = 53, then 53 – 6 = 47. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 26 and preserves the subtraction direction.
D7. 64 – 22 = 42. One route is 64 – 20 = 44, then 44 – 2 = 42. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 22 and preserves the subtraction direction.
D8. 91 – 37 = 54. One route is 91 – 30 = 61, then 61 – 7 = 54. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 37 and preserves the subtraction direction.
D9. 88 – 35 = 53. One route is 88 – 30 = 58, then 58 – 5 = 53. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 35 and preserves the subtraction direction.
D10. 76 – 28 = 48. One route is 76 – 20 = 56, then 56 – 8 = 48. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 28 and preserves the subtraction direction.
D11. 63 – 28 = 35. One route is 63 – 20 = 43, then 43 – 8 = 35. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 28 and preserves the subtraction direction.
D12. 72 – 39 = 33. One route is 72 – 30 = 42, then 42 – 9 = 33. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 39 and preserves the subtraction direction.
D13. 84 – 46 = 38. One route is 84 – 40 = 44, then 44 – 6 = 38. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 46 and preserves the subtraction direction.
D14. 93 – 57 = 36. One route is 93 – 50 = 43, then 43 – 7 = 36. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 57 and preserves the subtraction direction.
D15. 71 – 38 = 33. One route is 71 – 30 = 41, then 41 – 8 = 33. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 38 and preserves the subtraction direction.
D16. 65 – 27 = 38. One route is 65 – 20 = 45, then 45 – 7 = 38. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 27 and preserves the subtraction direction.
D17. 92 – 48 = 44. One route is 92 – 40 = 52, then 52 – 8 = 44. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 48 and preserves the subtraction direction.
D18. 83 – 36 = 47. One route is 83 – 30 = 53, then 53 – 6 = 47. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 36 and preserves the subtraction direction.
D19. 75 – 29 = 46. One route is 75 – 20 = 55, then 55 – 9 = 46. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 29 and preserves the subtraction direction.
D20. 62 – 27 = 35. One route is 62 – 20 = 42, then 42 – 7 = 35. If the second step is awkward, invite a bridge route instead of insisting on tens-then-ones. The important diagnostic is whether the child keeps track of both parts of 27 and preserves the subtraction direction.
Lab E — Subtraction as difference: count up when numbers are close
E1. 63 – 58 = 5. Instead of taking 58 away from 63, measure the gap from 58 up to 63. A convenient first jump is 2, followed by 3. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E2. 72 – 68 = 4. Instead of taking 68 away from 72, measure the gap from 68 up to 72. A convenient first jump is 2, followed by 2. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E3. 51 – 47 = 4. Instead of taking 47 away from 51, measure the gap from 47 up to 51. A convenient first jump is 3, followed by 1. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E4. 84 – 79 = 5. Instead of taking 79 away from 84, measure the gap from 79 up to 84. A convenient first jump is 1, followed by 4. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E5. 90 – 86 = 4. Instead of taking 86 away from 90, measure the gap from 86 up to 90. A convenient first jump is 4. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E6. 62 – 57 = 5. Instead of taking 57 away from 62, measure the gap from 57 up to 62. A convenient first jump is 3, followed by 2. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E7. 75 – 69 = 6. Instead of taking 69 away from 75, measure the gap from 69 up to 75. A convenient first jump is 1, followed by 5. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E8. 43 – 38 = 5. Instead of taking 38 away from 43, measure the gap from 38 up to 43. A convenient first jump is 2, followed by 3. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E9. 81 – 76 = 5. Instead of taking 76 away from 81, measure the gap from 76 up to 81. A convenient first jump is 4, followed by 1. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E10. 70 – 64 = 6. Instead of taking 64 away from 70, measure the gap from 64 up to 70. A convenient first jump is 6. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E11. 52 – 48 = 4. Instead of taking 48 away from 52, measure the gap from 48 up to 52. A convenient first jump is 2, followed by 2. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E12. 96 – 89 = 7. Instead of taking 89 away from 96, measure the gap from 89 up to 96. A convenient first jump is 1, followed by 6. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E13. 67 – 59 = 8. Instead of taking 59 away from 67, measure the gap from 59 up to 67. A convenient first jump is 1, followed by 7. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E14. 88 – 79 = 9. Instead of taking 79 away from 88, measure the gap from 79 up to 88. A convenient first jump is 1, followed by 8. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
E15. 61 – 54 = 7. Instead of taking 54 away from 61, measure the gap from 54 up to 61. A convenient first jump is 6, followed by 1. This route highlights subtraction as difference. It is especially efficient when the two numbers are close and helps children see that subtraction has more than one meaning.
Lab F — Compensation in subtraction
F1. 52 – 19 = 33. Replace 19 with the friendlier 20. 52 – 20 = 32. Because subtracting 20 removed 1 more than the original problem required, add 1 back: 32 + 1 = 33. The correction direction is the lesson: removing too much means restoring the excess.
F2. 73 – 29 = 44. Replace 29 with the friendlier 30. 73 – 30 = 43. Because subtracting 30 removed 1 more than the original problem required, add 1 back: 43 + 1 = 44. The correction direction is the lesson: removing too much means restoring the excess.
F3. 64 – 39 = 25. Replace 39 with the friendlier 40. 64 – 40 = 24. Because subtracting 40 removed 1 more than the original problem required, add 1 back: 24 + 1 = 25. The correction direction is the lesson: removing too much means restoring the excess.
F4. 85 – 49 = 36. Replace 49 with the friendlier 50. 85 – 50 = 35. Because subtracting 50 removed 1 more than the original problem required, add 1 back: 35 + 1 = 36. The correction direction is the lesson: removing too much means restoring the excess.
F5. 91 – 59 = 32. Replace 59 with the friendlier 60. 91 – 60 = 31. Because subtracting 60 removed 1 more than the original problem required, add 1 back: 31 + 1 = 32. The correction direction is the lesson: removing too much means restoring the excess.
F6. 72 – 38 = 34. Replace 38 with the friendlier 40. 72 – 40 = 32. Because subtracting 40 removed 2 more than the original problem required, add 2 back: 32 + 2 = 34. The correction direction is the lesson: removing too much means restoring the excess.
F7. 63 – 29 = 34. Replace 29 with the friendlier 30. 63 – 30 = 33. Because subtracting 30 removed 1 more than the original problem required, add 1 back: 33 + 1 = 34. The correction direction is the lesson: removing too much means restoring the excess.
F8. 84 – 58 = 26. Replace 58 with the friendlier 60. 84 – 60 = 24. Because subtracting 60 removed 2 more than the original problem required, add 2 back: 24 + 2 = 26. The correction direction is the lesson: removing too much means restoring the excess.
F9. 95 – 69 = 26. Replace 69 with the friendlier 70. 95 – 70 = 25. Because subtracting 70 removed 1 more than the original problem required, add 1 back: 25 + 1 = 26. The correction direction is the lesson: removing too much means restoring the excess.
F10. 61 – 28 = 33. Replace 28 with the friendlier 30. 61 – 30 = 31. Because subtracting 30 removed 2 more than the original problem required, add 2 back: 31 + 2 = 33. The correction direction is the lesson: removing too much means restoring the excess.
F11. 82 – 47 = 35. Replace 47 with the friendlier 50. 82 – 50 = 32. Because subtracting 50 removed 3 more than the original problem required, add 3 back: 32 + 3 = 35. The correction direction is the lesson: removing too much means restoring the excess.
F12. 74 – 59 = 15. Replace 59 with the friendlier 60. 74 – 60 = 14. Because subtracting 60 removed 1 more than the original problem required, add 1 back: 14 + 1 = 15. The correction direction is the lesson: removing too much means restoring the excess.
F13. 93 – 48 = 45. Replace 48 with the friendlier 50. 93 – 50 = 43. Because subtracting 50 removed 2 more than the original problem required, add 2 back: 43 + 2 = 45. The correction direction is the lesson: removing too much means restoring the excess.
F14. 71 – 39 = 32. Replace 39 with the friendlier 40. 71 – 40 = 31. Because subtracting 40 removed 1 more than the original problem required, add 1 back: 31 + 1 = 32. The correction direction is the lesson: removing too much means restoring the excess.
F15. 86 – 57 = 29. Replace 57 with the friendlier 60. 86 – 60 = 26. Because subtracting 60 removed 3 more than the original problem required, add 3 back: 26 + 3 = 29. The correction direction is the lesson: removing too much means restoring the excess.
Mixed Strategy Lab: Do Not Name the Method First
M1. 347 + 20 = 367. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M2. 426 + 30 = 456. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M3. 582 + 100 = 682. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M4. 731 + 200 = 931. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M5. 568 + 7 = 575. Compensation or a bridge to the next ten may be efficient because the first number is close to a friendly benchmark. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M6. 399 + 20 = 419. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M7. 248 + 50 = 298. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M8. 675 + 8 = 683. Several routes are plausible. The child should choose one and defend the choice from the number structure. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M9. 487 – 20 = 467. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M10. 625 – 100 = 525. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M11. 743 – 30 = 713. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M12. 900 – 200 = 700. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M13. 652 – 9 = 643. Compensation may help, but count-up or partitioning may be equally sensible depending on the gap. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M14. 501 – 20 = 481. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M15. 830 – 50 = 780. Place-value reasoning is likely the shortest route because the second quantity is already a whole group of tens or hundreds. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M16. 704 – 8 = 696. Compensation may help, but count-up or partitioning may be equally sensible depending on the gap. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M17. 39 + 27 = 66. Compensation or a bridge to the next ten may be efficient because the first number is close to a friendly benchmark. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M18. 58 + 26 = 84. Compensation or a bridge to the next ten may be efficient because the first number is close to a friendly benchmark. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M19. 49 + 35 = 84. Compensation or a bridge to the next ten may be efficient because the first number is close to a friendly benchmark. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M20. 63 – 28 = 35. Compensation may help, but count-up or partitioning may be equally sensible depending on the gap. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M21. 72 – 39 = 33. Compensation may help, but count-up or partitioning may be equally sensible depending on the gap. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M22. 52 – 19 = 33. Compensation may help, but count-up or partitioning may be equally sensible depending on the gap. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M23. 88 + 17 = 105. Compensation or a bridge to the next ten may be efficient because the first number is close to a friendly benchmark. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
M24. 61 – 54 = 7. Several routes are plausible. The child should choose one and defend the choice from the number structure. After solving, ask for a second correct route and one quick reason the answer is plausible. The goal is strategy selection without a chapter label telling the child what to do.
A Parent Diagnostic Conversation
The fastest way to learn what a child understands is often a short, calm conversation around one problem. Choose a question that is slightly revealing but not intimidating. Ask the child to try it without coaching. Watch the first move. Then ask for an explanation only after the attempt.
Suppose the problem is 48 + 27. If the child counts by ones, ask whether there is a way to make 50. If the child says 48 + 2 = 50 but then forgets that 25 remains, the partition is not being held reliably. If the child reaches 75 through 50 + 25, ask why moving 2 from 27 did not change the total. If the child can explain that the 2 was not added twice but transferred from one addend to the other, the structural understanding is strong.
Then change the numbers: 58 + 27. Do not repeat the hint. The changed question tests transfer. Return a few days later with 68 + 27. The delayed question tests retrieval. Finally place the strategy inside a money or story context. The context test shows whether the child can recognise the same relationship without the visual cue of a worksheet section title.
This four-part sequence—attempt, change, delay, context—gives more information than a single percentage score. It reveals whether the child can perform, transfer, retrieve and recognise.
How to Use Practice Without Creating Worksheet Dependence
Practice is necessary, but practice design changes what the child learns. Twenty questions with the same structure may improve immediate speed while giving the child a strong clue about which method to use. A mixed set is harder because the child must recognise the structure before calculating. Both have value, but they serve different phases.
Early practice should reduce unnecessary difficulty so a new relationship can stabilise. Once the child is accurate, vary the numbers. Then mix strategy families. Then remove method labels. Then revisit after delay. Then place the structure inside word problems, money or time. This gradually shifts responsibility from the worksheet to the learner.
A useful ratio for a ten-question set might be four questions from the current strategy, three from older strategies, two mixed questions and one explanation question. The exact ratio is not sacred. The principle is that retrieval and choice need to appear before an examination forces them to appear.
Correction should also be selective. Do not erase every mistake immediately. Ask the child to locate the first line where the mathematics stopped matching the original problem. This turns correction into diagnosis. The aim is not a clean page; it is a learner who can increasingly detect and repair a broken route.
What Progress Looks Like Before Marks Move
Parents understandably look at school marks, but many important improvements happen before the next formal assessment. A child may begin to start questions with less hesitation, use tens and hundreds more naturally, rely less on finger counting, choose a shorter route, catch an implausible answer, or explain an error more precisely. These are leading indicators.
Another strong indicator is a smaller hint. At first the teacher may need to say, “Can you make the next ten?” Later the hint becomes, “What do you notice?” Eventually the child notices independently. The shrinking size of the hint is evidence that responsibility is moving from adult to learner.
Delayed retrieval is equally important. If the child can use partitioning only on the day it was taught, the learning is still fragile. If the child can return after a week, recognise a suitable structure and choose the method without prompting, the strategy is becoming part of the learner’s working repertoire.
| Early improvement signal | Why it matters |
|---|---|
| Fewer one-by-one counts | Larger number units are becoming available. |
| More accurate place-value language | The child is connecting digits to quantities. |
| Can give two routes | Strategy range is expanding. |
| Chooses a route for a reason | Fluency is becoming flexible rather than mechanical. |
| Explains compensation direction | Equivalence is understood, not memorised. |
| Checks with estimation | Verification is becoming habitual. |
| Old strategies return after delay | Retrieval is strengthening. |
| Needs smaller hints | Independence is increasing. |
| Transfers to money/time/word problems | The idea is no longer trapped in a worksheet format. |
Alicia, Tricia and Kai Kai: One Afternoon, Three Ways to See a Number
Alicia sees 49 + 26 and immediately says, “I want 50.” She moves one from 26 to 49 and gets 50 + 25. Her strength is benchmark recognition. Tricia prefers to keep the second number intact as tens and ones: 49 + 20 = 69, then +6 = 75. Her strength is stable sequencing. Kai Kai writes 50 + 26 – 1. His strength is compensation language. All three reach 75.
The important classroom event is what happens next. The teacher asks, “Which method would still feel easy for 49 + 76?” Alicia’s transfer still works. Tricia’s tens-and-ones route also works. Kai Kai’s compensation remains efficient. Then the teacher changes the first number to 42. Alicia notices that moving to 50 now requires eight, which may no longer be the simplest route. Strategy choice has become contextual.
Later they meet 63 – 58. Tricia starts to subtract tens and ones, then stops because the route feels awkward. Alicia says the numbers are close and counts up five. Kai Kai draws two points on a number line and sees the gap. The class has moved beyond “the method for subtraction”. It is building a repertoire.
That repertoire matters because future Mathematics rarely announces the method in advance. A mixed problem set, a PSLE paper or a secondary-school algebra question asks the learner to recognise structure. The recognition habit begins much earlier than the examination.
From Primary 2 to Primary 3: What Should Be Stable?
Primary 3 increases the amount of structure a child must coordinate. Multiplication and division expand, fractions become more demanding, measurement and geometry develop, and word problems require stronger representation. The child does not need to be perfect at the end of Primary 2, but several foundations should be reasonably stable.
Whole-number place value should feel meaningful rather than ceremonial. Addition and subtraction should no longer depend entirely on counting by ones. The child should know the required P2 multiplication and division facts with growing fluency and understand the inverse relationship. Basic fraction language should be connected to equal parts. Money, measurement and time answers should retain units and context. Most importantly, the child should be increasingly willing to choose and check a method.
The next guide in this Bukit Timah Mathematics sequence is Primary 3 Mathematics | Multiplication and Division as One Reversible Structure. The previous foundation is Primary 1 Mathematics | Number Bonds as Reversible Relationships, Not Facts to Recite. Together they form a coherent progression from additive relationships into multiplicative structure.
Long-Range View: Why This Matters for PSLE Without Turning P2 Into PSLE Drilling
Parents often ask whether Primary 2 work should be designed for PSLE. The best answer is yes in principle and no in form. Yes, because strong place value, flexible calculation, representation, strategy selection and checking are reused throughout Primary Mathematics. No, because a seven- or eight-year-old does not need Primary 6 paper pressure or premature question formats to build those foundations.
The current PSLE Mathematics framework ultimately values straightforward computation, application across contexts and mathematical reasoning including strategy selection. The route to that capability is not six years of examination drilling. It is a sequence of increasingly powerful mathematical structures, practised until they can be retrieved and transferred.
When Primary 6 arrives, preparation should eventually shift toward mixed retrieval, paper control, timing and examination conversion. That later job is explained in PSLE Mathematics Tuition Bukit Timah | When to Stop Rebuilding and Start Exam Conversion. Primary 2 has a different job: build the number system well enough that later exam preparation has something reliable to operate on.
Frequently Asked Questions
Should my Primary 2 child memorise multiplication tables?
Yes, growing automatic recall is useful, but the facts should be attached to equal groups, arrays, repeated structure and inverse division relationships. Understanding and recall should reinforce each other.
Is finger counting always bad?
No. It is an early strategy. The concern is persistent dependence when more efficient number relationships are available. The goal is to expand the child’s units of thought, not shame an earlier method.
Should every sum be done mentally?
No. Mental methods, written algorithms, drawings and concrete models serve different purposes. A strong learner chooses a representation that fits the problem and can check the result.
What is the difference between partitioning and compensation?
Partitioning decomposes a quantity into parts that sum to the original quantity. Compensation deliberately changes a number to make the calculation friendlier, then repairs that change so the final expression remains equivalent.
Why does my child forget the last part after splitting a number?
The strategy may exceed working-memory capacity. Record the partition explicitly, use a number line, or choose a smaller-step route. The goal is reliable reasoning, not maximum mental load.
Is the bar model compulsory for every word problem?
No. A representation is useful when it clarifies a relationship. Simple questions may not need a full bar model. More complex comparison or part-whole structures may benefit greatly from one.
How much home practice is enough?
A short, focused routine can be sufficient when it targets one clear relationship, includes retrieval from older learning and stops before fatigue. Quality and spacing matter more than sheer page count.
Should I teach Primary 3 topics early if P2 work is easy?
Only after current learning is stable, transferable and not creating unnecessary workload. Enrichment can deepen explanation, strategy comparison, puzzles and applications instead of merely racing ahead in the syllabus.
What if my child gets correct answers but uses a long method?
That is a useful starting point. Ask for a second method and compare the number of steps. Strategy efficiency can be taught without invalidating a correct route.
What if my child is very fast but cannot explain anything?
Sample the reasoning. Speed from secure facts is valuable. Speed from guessing or opaque procedures is fragile. A few explanation checks can distinguish the two.
What if my child understands at home but fails school worksheets?
Compare the contexts. The child may depend on hints, may not recognise the structure when the worksheet is mixed, or may lose attention under independent conditions. Test changed and delayed questions.
What if my child is anxious about Mathematics?
Reduce public speed pressure, use manageable questions, make error correction specific rather than personal, and create enough successful independent attempts for the learner to regain control.
Does Primary 2 performance predict a future PSLE grade?
A single P2 score should not be treated as a deterministic forecast of a later PSLE result. Development, teaching, practice, health, school context and many other factors change over several years. Use current evidence to decide the next useful learning action.
Is small-group tuition better than one-to-one?
Neither format is automatically better. One-to-one offers maximum individual attention; a well-run small group can add peer strategy comparison and independent work periods. The fit depends on the learner, teacher and class design.
How do I know whether tuition is helping?
Look for smaller hints, better retrieval after delay, stronger strategy choice, fewer repeated errors, improved explanation and transfer to school work. Marks matter, but process indicators often move first.
What is a good mental-maths question for the dinner table?
Ask something with visible structure, such as 49 + 27, and then ask, “What made your route easy?” Avoid turning every family interaction into a quiz.
Why teach estimation if the child can calculate exactly?
Estimation creates a plausibility check. It helps children detect answers that are impossible in scale and encourages attention to magnitude rather than symbol manipulation alone.
Should my child always use the teacher’s method?
The child should understand school-taught methods and be able to communicate clearly. Alternative correct strategies can also be valuable when they are mathematically sound, efficient and explainable.
What does ‘mental calculation’ actually mean?
It means carrying out all or most of a calculation without a full written algorithm, often by using number relationships, place value, partitioning, compensation, known facts and estimation. A child may still jot a helpful intermediate number.
Where should we go next?
Use the Mathematics World for the broad learning route, the Mathematics Article Directory for topic guides, and the Primary 3 page for the next year-level bridge.
Reference Shelf and Further Reading
- Ministry of Education Singapore — Primary Mathematics Syllabus, updated October 2025
- Singapore Examinations and Assessment Board — 2026 PSLE formats
- ERIC — Teaching Mental Computation Strategies in Early Mathematics
- ERIC — A Proposed Framework for Examining Basic Number Sense
- ERIC — Number Sense on the Number Line
- eduKateSingapore Mathematics World
- eduKateSingapore Mathematics Article Directory
- Bukit Timah Tutor — Primary 2 Mathematics specialist programme route
Final Principle
Primary 2 mental calculation is successful when a child can change the route without changing the mathematics.
Partitioning teaches that a number can be decomposed and recomposed. Compensation teaches that a temporary change must be repaired. Place value explains why tens and hundreds can be treated as coherent units. Number bonds provide bridges to friendly benchmarks. Estimation checks magnitude. Word problems test whether the structure can be recognised outside a naked calculation. Explanation makes the reasoning visible. Together, these are not disconnected techniques. They are one early lesson in mathematical equivalence.
The aim is not to manufacture a child who performs mental arithmetic for display. It is to develop a learner who sees structure, selects a sensible method, preserves meaning, notices when something has gone wrong and can recover. That is a Primary 2 capability. It is also the beginning of much more advanced mathematics.
Appendix A — 60 Diagnostic Micro-Cases for Parents and Teachers
Case 1: 47 + 6. Observation: Child counts six single steps. Diagnostic probe: Can the child use 3 to reach 50, then add the remaining 3? Purpose: Tests complements and bridging. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 2: 58 + 7. Observation: Child says 66. Diagnostic probe: Ask the child to show the jump to 60 first. Purpose: Tests tracking of the remaining addend. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 3: 39 + 18. Observation: Child answers 58. Diagnostic probe: Ask what changed when 39 became 40. Purpose: Tests omitted compensation. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 4: 52 – 19. Observation: Child answers 31. Diagnostic probe: Ask whether subtracting 20 removed too much or too little. Purpose: Tests compensation direction. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 5: 347 + 20. Observation: Child writes 349. Diagnostic probe: Build 347 as hundreds, tens and ones. Purpose: Tests place-value interpretation. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 6: 426 + 100. Observation: Child changes the tens digit. Diagnostic probe: Ask which place represents hundreds. Purpose: Tests positional value. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 7: 63 – 58. Observation: Child begins a long take-away route. Diagnostic probe: Ask whether measuring the gap would be simpler. Purpose: Tests subtraction as difference. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 8: 74 – 32. Observation: Child loses the 2 after subtracting 30. Diagnostic probe: Write 32 = 30 + 2 before calculating. Purpose: Tests partition retention. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 9: 49 + 36. Observation: Child uses vertical algorithm correctly. Diagnostic probe: Ask for a mental route through 50. Purpose: Tests flexibility rather than correctness. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 10: 68 + 25. Observation: Child uses 68 + 20 + 5. Diagnostic probe: Ask for another route through 70. Purpose: Tests strategy range. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 11: 83 – 29. Observation: Child subtracts 30 then subtracts 1. Diagnostic probe: Ask whether the temporary subtraction removed too much. Purpose: Tests correction sign. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 12: 91 – 57. Observation: Child is accurate but very slow. Diagnostic probe: Ask whether 57 can be partitioned around a friendly ten. Purpose: Tests efficiency. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 13: 5 × 4. Observation: Child chants answer without model. Diagnostic probe: Ask for five groups of four using objects or a drawing. Purpose: Tests fact meaning. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 14: 20 ÷ 5. Observation: Child answers 4. Diagnostic probe: Ask for both grouping and sharing stories. Purpose: Tests division interpretation. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 15: 3 × 4 and 12 ÷ 4. Observation: Child treats them as unrelated. Diagnostic probe: Build one array and write all related facts. Purpose: Tests inverse relationship. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 16: $3.00 – $2.70. Observation: Child writes 0.30 but cannot say cents. Diagnostic probe: Ask for the amount as cents and as dollars/cents. Purpose: Tests money meaning and notation. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 17: 2:35 plus 20 minutes. Observation: Child changes hour too early. Diagnostic probe: Bridge to 3:00 and compare 20 with 25 minutes. Purpose: Tests time structure. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 18: Picture graph: one icon = 2. Observation: Child counts icons as one each. Diagnostic probe: Ask what one icon stands for before counting. Purpose: Tests scaled units. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 19: One quarter vs one third. Observation: Child thinks 4 is larger so one quarter is larger. Diagnostic probe: Use equal wholes cut into different numbers of equal parts. Purpose: Tests denominator meaning. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 20: 2/5 + 1/5. Observation: Child adds denominators. Diagnostic probe: Show five equal parts and count selected parts. Purpose: Tests like-fraction structure. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 21: 38 + 7. Observation: Child counts fingers. Diagnostic probe: Ask what 38 needs to reach 40. Purpose: Tests benchmark access. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 22: 59 + 4. Observation: Child answers 62. Diagnostic probe: Ask for 59 + 1 first. Purpose: Tests partition of 4 into 1 and 3. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 23: 48 + 37. Observation: Child reaches 50 then adds all 37. Diagnostic probe: Ask where the 2 came from. Purpose: Tests double-counting after transfer. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 24: 67 + 26. Observation: Child says 67 + 20 = 87, then +6 = 92. Diagnostic probe: Ask for a quick estimate around 70 + 30. Purpose: Tests arithmetic plus estimation. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 25: 95 – 43. Observation: Child answers 58. Diagnostic probe: Ask for 95 – 40 first and record it. Purpose: Tests subtraction fact accuracy. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 26: 72 – 39. Observation: Child freezes. Diagnostic probe: Offer two routes: subtract 40 then add 1, or count up from 39. Purpose: Tests alternative representation. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 27: 88 + 17. Observation: Child says 105. Diagnostic probe: Ask for estimation around 90 + 20. Purpose: Tests plausibility checking. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 28: 71 – 38. Observation: Child answers 43. Diagnostic probe: Ask whether 70 – 40 would be near 30 or 40. Purpose: Tests magnitude sense. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 29: 582 + 100. Observation: Child writes 592. Diagnostic probe: Use place-value blocks or expanded notation. Purpose: Tests hundreds addition. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 30: 900 – 200. Observation: Child counts backwards by tens. Diagnostic probe: Ask how many hundreds are being removed. Purpose: Tests unit size. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 31: 347 + 6. Observation: Child writes full algorithm. Diagnostic probe: Ask for a mental bridge to 350. Purpose: Tests strategy selection. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 32: 501 – 20. Observation: Child becomes confused by zero. Diagnostic probe: Represent 501 as 50 tens and 1 one if useful. Purpose: Tests regrouped place-value representations. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 33: 28 + 45. Observation: Child uses 20 + 40 and 8 + 5, then forgets to combine. Diagnostic probe: Record partial sums explicitly. Purpose: Tests recomposition. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 34: 36 + 29. Observation: Child adds 30 but forgets correction. Diagnostic probe: Ask whether 29 is one less or one more than 30. Purpose: Tests compensation relation. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 35: 84 – 58. Observation: Child loses track across the ten. Diagnostic probe: Try counting up 58→60→80→84. Purpose: Tests difference route. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 36: 96 – 89. Observation: Child attempts a lengthy algorithm. Diagnostic probe: Ask for the gap between 89 and 96. Purpose: Tests near-number subtraction. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 37: 42 + 25. Observation: Child insists compensation is required. Diagnostic probe: Compare number of steps with tens-and-ones partitioning. Purpose: Tests method appropriateness. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 38: 99 + 17. Observation: Child does not notice 100. Diagnostic probe: Ask what 99 needs to reach 100. Purpose: Tests benchmark recognition. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 39: 70 – 64. Observation: Child counts down six steps. Diagnostic probe: Ask whether counting up from 64 is easier. Purpose: Tests directional choice. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 40: 61 – 54. Observation: Child says 17. Diagnostic probe: Ask for an estimate and then measure the gap. Purpose: Tests magnitude and difference. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 41: 4 × 5. Observation: Child knows 20; 5 × 4 feels new. Diagnostic probe: Use an array and rotate it. Purpose: Tests commutative structure at an intuitive level. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 42: 15 ÷ 3. Observation: Child counts randomly. Diagnostic probe: Build 15 objects and make groups of 3. Purpose: Tests grouping interpretation. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 43: 12 ÷ 4. Observation: Child makes four groups of four. Diagnostic probe: Count the total and compare with 12. Purpose: Tests grouping vs sharing control. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 44: 3 groups of 5. Observation: Child writes 3 + 5. Diagnostic probe: Act out equal groups and compare addition sentence with multiplication sentence. Purpose: Tests operation meaning. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 45: $1.50 + 40 cents. Observation: Child writes $1.90 but cannot justify. Diagnostic probe: Represent $1.50 as 150 cents, then convert back. Purpose: Tests notation equivalence. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 46: 1 hour 20 minutes. Observation: Child says 120 minutes. Diagnostic probe: Build 60 + 20. Purpose: Tests time-unit conversion reasoning. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 47: A 1 kg bag and a 500 g bag. Observation: Child adds 1 + 500. Diagnostic probe: Ask what quantities the numerals represent and whether units match. Purpose: Tests unit awareness. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 48: 2 litres and 1 litre. Observation: Child writes 3 without unit. Diagnostic probe: Ask what the 3 counts. Purpose: Tests number-with-unit discipline. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 49: Cube vs cuboid. Observation: Child classifies only by size. Diagnostic probe: Ask about faces and shape properties rather than visual scale. Purpose: Tests attribute-based classification. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 50: Pattern alternates shape and colour. Observation: Child follows only colour. Diagnostic probe: Ask which attributes are changing. Purpose: Tests multi-attribute pattern reading. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 51: Picture graph scale 5. Observation: Child totals pictures but ignores scale. Diagnostic probe: Write repeated groups of 5 under each icon. Purpose: Tests multiplicative interpretation. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 52: 47 + 26 after a week. Observation: Child cannot remember the taught route. Diagnostic probe: Do not reteach immediately; ask what parts of 26 are visible. Purpose: Tests retrieval and reconstruction. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 53: 39 + 18 in a shopping story. Observation: Child no longer recognises compensation. Diagnostic probe: Compare the contextual quantities with the naked calculation. Purpose: Tests transfer across representation. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 54: 58 + 27 mixed among subtraction questions. Observation: Child waits for chapter cue. Diagnostic probe: Ask what the operation sign and numbers suggest. Purpose: Tests independent recognition. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 55: 63 – 58 after seeing 63 – 28. Observation: Child uses the same partition method mechanically. Diagnostic probe: Ask whether closeness changes the efficient route. Purpose: Tests contextual flexibility. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 56: 347 + 30 after practising 347 + 3. Observation: Child treats 30 as 3. Diagnostic probe: Ask what unit the 3 represents in 30. Purpose: Tests scaling by place value. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 57: One correct test score spike. Observation: Parent assumes mastery. Diagnostic probe: Check a delayed, changed and contextual version. Purpose: Tests stability beyond one performance. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 58: Repeated correction, same error. Observation: Child copies worked solutions. Diagnostic probe: Ask the child to generate the first step from a blank page. Purpose: Tests productive retrieval. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 59: Fast answers, frequent unexplained slips. Observation: Child may prioritise speed over verification. Diagnostic probe: Add an estimate-and-check requirement. Purpose: Tests control under fluency. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Case 60: Slow but accurate reasoning. Observation: Parent worries about speed. Diagnostic probe: Stabilise structure first; then build fact fluency and efficient selection. Purpose: Tests developmental sequencing. Do not turn the probe into a lecture; give the child enough space to reveal the current model. The response to one changed question is often more informative than ten repetitions of the original.
Appendix B — 40 Strategy-Comparison Prompts
Prompt 1: 46 + 27. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 2: 58 + 34. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 3: 69 + 16. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 4: 38 + 45. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 5: 57 + 28. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 6: 49 + 37. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 7: 68 + 24. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 8: 29 + 56. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 9: 77 + 18. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 10: 59 + 33. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 11: 73 – 28. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 12: 84 – 39. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 13: 92 – 47. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 14: 65 – 29. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 15: 81 – 36. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 16: 74 – 58. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 17: 63 – 27. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 18: 90 – 46. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 19: 72 – 19. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 20: 86 – 57. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 21: 347 + 20. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 22: 426 + 30. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 23: 582 + 100. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 24: 731 + 200. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 25: 568 + 7. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 26: 675 + 8. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 27: 487 – 20. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 28: 625 – 100. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 29: 743 – 30. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 30: 900 – 200. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 31: 39 + 18. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 32: 52 – 19. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 33: 99 + 27. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 34: 83 – 29. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 35: 48 + 7. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 36: 63 – 58. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 37: 72 – 68. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 38: 88 + 17. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 39: 61 – 54. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Prompt 40: 98 + 23. Solve it once. Then find a second correct route. Compare the routes using three criteria: number of mental steps, risk of losing a quantity, and ease of checking. Finally, change one number slightly and decide whether your preferred method still makes sense. This comparison turns fluency into strategy choice rather than method loyalty.
Appendix C — Teacher and Parent Observation Rubric
| Dimension | Emerging | Developing | Secure | Extending |
|---|---|---|---|---|
| Place value | Treats digits as isolated symbols | Can name places but may misapply tens/hundreds in operations | Uses hundreds/tens/ones reliably in calculation | Recomposes numbers flexibly for efficient strategies |
| Number bonds | Relies heavily on counting | Recalls some bonds but does not always use them | Uses complements to bridge through tens | Uses benchmarks flexibly across larger numbers |
| Partitioning | May lose a part | Can split tens/ones with support | Partitions accurately and recomposes reliably | Chooses non-standard partitions for efficiency |
| Compensation | Rounds without repairing | Repairs with reminders | Explains direction of correction | Selects compensation only when it improves the route |
| Subtraction meaning | Mostly take-away | Can use count-up with prompting | Chooses between take-away and difference | Explains why different models are equivalent |
| Multiplication/division | Facts are isolated or counted | Understands equal groups but recall is uneven | Connects multiplication and division facts | Uses known facts to derive unfamiliar related facts |
| Word problems | Chooses operation from keywords | Can represent simple relationships with support | Identifies quantities and relationships before calculating | Compares multiple representations and selects the clearest |
| Checking | Rarely checks | Checks when prompted | Uses estimation or inverse reasoning independently | Chooses checks strategically based on the problem |
| Explanation | Gives answer only | Can describe steps | Explains why steps preserve meaning | Compares methods and critiques efficiency |
| Independence | Waits for adult cues | Begins with small prompts | Starts and checks most tasks independently | Can diagnose and repair own errors |
Appendix D — A 30-Day Micro-Practice Calendar
Day 1. Make numbers to 1000 with hundreds, tens and ones; decompose five numbers in two ways. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 2. Practise complements to 10 and next-ten bridges in context. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 3. Add multiples of 10 to three-digit numbers mentally. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 4. Subtract multiples of 10 from three-digit numbers mentally. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 5. Mixed retrieval from Days 1–4; no method labels. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 6. Partition two-digit addition into tens and ones. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 7. Bridge through the next ten for five carefully chosen additions. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 8. Compare partitioning and bridge strategies on the same three problems. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 9. Partition subtraction; record both parts before calculating. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 10. Use count-up for close-number subtraction. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 11. Compensation in addition with numbers ending in 9. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 12. Compensation in subtraction with subtrahends ending in 9. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 13. Mixed strategy choice: eight questions, explain two. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 14. Rest or use a light game involving complements and place value. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 15. Multiplication as equal groups for 2, 5 and 10 facts. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 16. Multiplication and division fact families for 3 and 4. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 17. Arrays and repeated groups; write related equations. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 18. One-step multiplication/division word problems. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 19. Money: totals and change using benchmark amounts. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 20. Time: bridge to the next hour in simple duration tasks. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 21. Measurement and units: estimate first, then calculate or compare. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 22. Fractions: build equal-part models and compare unit fractions. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 23. Like fractions within one whole using pictures before symbols. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 24. Picture graphs with scales; say what one icon represents before counting. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 25. Mixed retrieval from the previous ten days. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 26. Word problems: identify quantities and relationship before any arithmetic. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 27. Solve one problem three ways and compare the routes. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 28. Error hunt: inspect four incorrect worked solutions and locate the first broken step. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 29. Delayed check: redo selected Day 6, 11, 16 and 21 ideas without notes. Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.
Day 30. Reflection: which strategies are now automatic, which still need prompts, and what should be practised next? Keep the session brief enough that attention remains high. End with one question the child can solve independently so the final experience is one of control rather than exhaustion.