Ask two learners to calculate 49 + 36 mentally.
One tries to imagine the vertical algorithm:
9 + 6 is 15, carry 1, 4 + 3 + 1 …
The other thinks:
49 is one away from 50.
Take one from 36.
50 + 35 = 85.
Both learners may reach the same answer.
But they are doing different kinds of mathematics.
Mental calculation is not the written algorithm with the paper removed. It is the ability to reorganise numbers around useful structure while keeping the value correct.
This distinction matters because mental arithmetic is one of the first places where mathematical flexibility becomes visible.
A learner with only one method may still be accurate.
A learner with several connected methods can choose between them.
That choice becomes increasingly important as numbers grow, problems become multi-step and written working has to be reserved for the parts that genuinely need it.
The current Singapore Primary Mathematics syllabus explicitly includes mental calculation from Primary 1 onward. Primary 2 extends mental addition and subtraction to three-digit numbers with ones, tens or hundreds. Two-digit mental calculation is therefore not a separate syllabus island. It is a critical bridge between early number bonds and later place-value fluency.
The quick answer: what makes a mental strategy good?
A good mental strategy should do at least three things:
- preserve the value of the problem;
- reduce the cognitive load compared with a less suitable method;
- make use of a mathematical relationship the learner understands.
There is no single best strategy for every two-digit calculation.
For 49 + 36, compensation is attractive.
For 42 + 30, place-value addition is obvious.
For 27 + 27, doubling may be fastest.
For 61 − 59, finding the difference may be easier than formal subtraction.
For 68 + 7, bridging through 70 may be natural.
The core question is:
What structure in these particular numbers can make the calculation simpler?
Strategy 1: add tens and ones by place value
Take 34 + 23.
Partition:
34 = 30 + 4.
23 = 20 + 3.
Combine like place values:
30 + 20 = 50.
4 + 3 = 7.
50 + 7 = 57.
This strategy is especially useful when no regrouping is needed.
It strengthens the understanding that a two-digit number is composed of tens and ones, not an indivisible numeral.
Strategy 2: add the tens, then the ones
For 46 + 32:
46 + 30 = 76.
76 + 2 = 78.
This preserves one addend intact while decomposing the other.
Some learners find this easier than splitting both numbers because fewer partial quantities must be held in working memory.
The mathematics is still place-value decomposition:
32 = 30 + 2.
Then:
46 + 32 = 46 + 30 + 2.
Strategy 3: bridge through the next ten
For 58 + 7:
58 needs 2 to make 60.
Split 7 into 2 and 5.
58 + 7 = 60 + 5 = 65.
This is the older Primary 1 make-ten idea operating at a larger scale.
The benchmark is no longer only 10.
It may be 20, 30, 40, 50, 60 or another nearby multiple of ten.
This is one of the clearest examples of mathematical transfer:
the strategy survives, but the scale changes.
Strategy 4: compensation in addition
Compensation changes one addend to make it friendlier, then adjusts another quantity so the total remains unchanged.
For 49 + 36:
Move 1 from 36 to 49.
49 + 36 = 50 + 35 = 85.
Why does this work?
One addend increased by 1 while the other decreased by 1.
The total change is zero.
The sum is invariant.
This is a much stronger explanation than “round 49 to 50 and remember to fix it somehow”.
Compensation is controlled equivalence.
Strategy 5: round one addend, then correct
A related strategy keeps the second addend unchanged at first.
For 39 + 26:
39 is one less than 40.
40 + 26 = 66.
Because we added one too much, subtract one:
66 − 1 = 65.
This is often easier to explain initially because the correction is explicit.
Compare it with equal compensation:
39 + 26 = 40 + 25 = 65.
Both are valid.
The learner should understand how the adjustment preserves the original total.
Strategy 6: doubles
For 27 + 27:
double 20 = 40.
double 7 = 14.
40 + 14 = 54.
Or:
27 × 2 = 54.
Doubles connect addition to multiplicative thinking.
A learner who recognises two equal addends has access to a compressed representation.
Strategy 7: near doubles
For 27 + 28:
double 27 = 54.
One more gives 55.
Or double 28 = 56, then subtract one.
This strategy depends on recognising closeness to an equal pair.
It is particularly efficient when the addends differ by one or two.
Strategy 8: subtract tens, then ones
For 74 − 32:
74 − 30 = 44.
44 − 2 = 42.
This strategy decomposes the subtrahend by place value.
It works especially cleanly when the intermediate calculation does not create awkward boundaries.
Again, one number remains intact while the other is partitioned.
Strategy 9: compensation in subtraction
For 63 − 29:
Subtract 30:
63 − 30 = 33.
But 30 is one more than 29, so we subtracted one too much.
Add one back:
33 + 1 = 34.
The correction must follow the direction of the error.
If we subtract too much, add back.
If we subtract too little, subtract the missing amount.
Strategy 10: find the difference by counting up
For 61 − 58, subtraction as removal is cumbersome mentally.
Think instead:
58 + ? = 61.
58 → 60 is 2.
60 → 61 is 1.
Total difference = 3.
This uses subtraction as difference or missing addition rather than take-away.
It is often the best strategy when two numbers are close together.
Strategy 11: use a friendly intermediate number
For 83 − 27:
Subtract 3 first to reach 80:
83 − 3 = 80.
There are 24 left to subtract:
80 − 24 = 56.
Or decompose 27 as 20 + 7:
83 − 20 = 63.
63 − 7 = 56.
Both routes are valid.
Which is easier depends on the learner and the number relationships noticed.
Strategy 12: preserve the difference by shifting both numbers
Consider 72 − 38.
Add 2 to both numbers:
74 − 40.
The difference remains the same because both numbers moved equally on the number line.
74 − 40 = 34.
This is a sophisticated compensation strategy.
It should be introduced only when the learner can explain why equal shifts preserve difference.
In addition, equal-and-opposite adjustments can preserve the sum. In subtraction, equal adjustments to both numbers can preserve the difference.
Mental calculation depends on number bonds that have grown up
Primary 1 number bonds do not disappear.
They scale.
If the learner knows:
8 + 2 = 10,
then the same structure supports:
- 18 + 2 = 20;
- 28 + 2 = 30;
- 48 + 2 = 50;
- 98 + 2 = 100.
If 7 needs 3 to make 10, then:
- 47 needs 3 to make 50;
- 67 needs 3 to make 70.
Mental calculation becomes easier when the learner recognises old relationships at new scales.
Place value must stay visible mentally
Take 46 + 20.
A learner who understands place value thinks:
4 tens + 2 tens = 6 tens; ones remain 6.
Answer: 66.
A learner using superficial digit editing may also get 66, but the method becomes fragile at boundaries.
Try:
86 + 20 = 106.
Eight tens + two tens = ten tens = one hundred.
The boundary exposes whether the learner is thinking in units or editing digits by pattern.
A good mental strategy should be explainable
Fast answers can hide fragile reasoning.
After a mental calculation, ask:
“What did you do to the numbers?”
A strong explanation does not need formal terminology.
For 49 + 36:
“I moved one from 36 to 49 so I could make 50. Then I had 50 + 35.”
That explanation shows conservation of the sum.
For 63 − 29:
“I subtracted 30 because it was easier, but that was one too much, so I added one back.”
That explanation shows controlled correction.
Strategy choice should depend on the numbers
Consider these sums:
- 42 + 30;
- 49 + 36;
- 27 + 28;
- 58 + 7.
A rigid learner may apply the same method to all four.
A flexible learner sees different structures:
- 42 + 30 → add tens;
- 49 + 36 → compensate to 50;
- 27 + 28 → near double;
- 58 + 7 → bridge through 60.
This is not about showing off many tricks.
It is about reading the mathematical affordances of the numbers.
Common misconception 1: mental means no intermediate steps
A mental strategy may still involve several internal steps.
For 68 + 27:
68 + 20 = 88.
88 + 7 = 95.
There is nothing mathematically inferior about that sequence.
The aim is efficient reasoning without unnecessary written procedure, not instantaneous answer production.
Common misconception 2: mental calculation should never use fingers or jottings
Temporary external support can be useful while a strategy is being learned.
A learner might jot 50 + 35 while explaining 49 + 36.
That does not invalidate the strategy.
The long-term goal is for the key relationships to become available with less support.
Support should fade when it is no longer carrying essential thinking.
Common misconception 3: a fast strategy is always the best strategy
Efficiency includes reliability.
If a clever compensation route causes frequent sign errors, a slightly longer partitioning method may be better for that learner until the structure is more secure.
The best method is not the method with the fewest spoken words.
It is the method that is accurate, understandable and appropriately efficient for the numbers and the learner.
Common misconception 4: compensation means changing the problem freely
A child turns 49 + 36 into 50 + 36 and forgets to compensate.
The answer becomes one too large.
Compensation is valid only when the change is mathematically accounted for.
Either:
50 + 36 = 86, then subtract 1 → 85;
or move that 1 from the other addend:
50 + 35 = 85.
The adjustment must preserve equivalence.
Common misconception 5: subtraction compensation uses the same correction direction as addition
For 63 − 29, if we subtract 30 we have subtracted one too much.
So we add one back.
A learner may instead subtract another one, producing 32.
The fix is not memorising “plus for subtraction”.
Ask:
Did we remove too much or too little?
The direction of correction follows the meaning.
Common misconception 6: partition every number even when it makes the calculation harder
For 99 + 34, splitting both numbers into tens and ones works:
90 + 30 + 9 + 4.
But compensation is simpler:
100 + 33 = 133.
A strategy should simplify the structure, not merely demonstrate that a decomposition is possible.
Common misconception 7: mental methods do not need checking
Mental calculation is vulnerable to lost intermediate values.
Estimate before or after.
For 49 + 36:
50 + 40 is about 90.
An answer such as 185 is impossible.
Then verify with another exact method if uncertainty remains.
Mental does not mean uncheckable.
A strategy-selection diagnostic
Give four calculations without telling the learner which method to use:
- 43 + 20;
- 49 + 28;
- 26 + 27;
- 72 − 69.
Observe:
- Does the learner use one method for all four?
- Does the learner notice the friendly ten in 49?
- Does the learner see the near-double in 26 + 27?
- Does the learner find the small difference between 72 and 69?
- Can the child explain why the chosen strategy fits?
The final answer alone does not reveal strategy flexibility.
A progression from supported to independent mental calculation
Stage 1: make structure visible
Use number lines, base-ten materials, ten-frames or part–whole diagrams to show why a strategy works.
Stage 2: describe the transformation
The learner says what changed and what stayed the same.
Stage 3: use short jottings
Record only key intermediate values such as 50 + 35.
Stage 4: solve mentally and explain afterwards
The strategy is internalised enough that the learner need not externalise every step.
Stage 5: choose among mixed methods
Problems are mixed so no worksheet heading reveals the intended strategy.
This final stage tests method selection rather than imitation.
The transfer test: change the surface story
A learner who can calculate 49 + 36 may fail when the same sum appears in words.
Example:
A library shelf has 49 books. Another 36 books are added. How many books are there now?
The learner must first identify addition from the relationship.
Then the mental strategy can operate.
49 + 36 → 50 + 35 → 85.
This is a stronger test than another naked calculation because the method is no longer cued directly.
The reverse transfer: choose when not to use mental calculation
Strategy flexibility includes restraint.
For a long multi-step calculation, forcing everything mentally can increase error risk.
A learner should ask:
- Is there a clear mental structure?
- Can I hold the intermediate values reliably?
- Would a written method reduce risk?
- Do I need a record for later steps?
Mental calculation is a tool, not a badge of intelligence.
The best mathematician is not the person who refuses paper.
It is the person who chooses representations intelligently.
A two-digit mental calculation diagnostic ladder
Check 1: basic facts
Are number bonds, doubles and complements to ten sufficiently retrievable?
Check 2: place value
Can tens and ones be decomposed and recombined?
Check 3: friendly tens
Can the learner see how far a number is from the next multiple of ten?
Check 4: compensation
Can the learner adjust numbers while preserving sum or difference appropriately?
Check 5: doubles and near doubles
Can equal or nearly equal addends be recognised?
Check 6: subtraction meanings
Can the learner switch between take-away, difference and missing addition?
Check 7: method selection
Can the learner choose a strategy without being told which one?
Check 8: verification
Can the learner estimate or use another method to check?
This ladder separates speed from structural fluency.
What parents should listen for
- “I made 49 into 50 and moved one from the other number.”
- “I added the three tens first, then the two ones.”
- “Twenty-seven and twenty-eight are almost doubles.”
- “Sixty-one minus fifty-eight is just the gap from fifty-eight to sixty-one.”
- “I used the written method because there were too many intermediate steps to hold safely.”
The final sentence is especially mature. It shows that the learner is choosing a representation based on cognitive and mathematical demands.
What teachers and tutors should avoid
- Avoid teaching a catalogue of disconnected tricks. Connect every strategy to place value, number bonds or invariance.
- Avoid forcing one strategy on every number pair. Strategy should respond to structure.
- Avoid treating speed as the only evidence of fluency. Ask for explanation and transfer.
- Avoid banning jottings too early. Fade support when it is no longer needed.
- Avoid letting compensation become unaccounted number changing. Every adjustment must preserve the original value relation.
- Avoid making mental calculation compete with written algorithms. They are complementary tools.
How this fits Singapore Primary Mathematics
The current MOE Primary Mathematics syllabus includes mental calculation involving addition and subtraction from Primary 1. At Primary 2, the scope includes mental calculation involving a three-digit number and ones, tens or hundreds, alongside addition and subtraction algorithms up to three digits.
Two-digit mental strategies are therefore foundational rather than optional enrichment.
They help learners coordinate earlier number-bond knowledge with place value, while preparing for larger calculations in which adding 1, 10 or 100 should be understood structurally.
The broader syllabus framework emphasises mathematical problem solving, reasoning, communication and metacognition. Strategy choice belongs directly inside that aim.
How do we know flexible strategy use matters?
Evidence guidance in mathematics education supports building conceptual understanding together with procedural fluency, using representations and derived strategies rather than relying only on memorised procedures.
The Institute of Education Sciences’ mathematics intervention guidance includes derived fact strategies such as making ten and stresses connections among number relationships, visual representations and efficient calculation.
The Education Endowment Foundation’s mathematics guidance likewise emphasises representations, comparison, connections and mathematical thinking rather than treating fluency as mere repetition.
The educational conclusion should remain careful.
More strategies are not automatically better.
The useful goal is a small, connected repertoire that the learner understands well enough to choose intelligently.
A readiness checkpoint for flexible mental calculation
- Can the learner decompose two-digit numbers into tens and ones?
- Can the learner add or subtract whole tens mentally?
- Can the learner bridge through a nearby ten?
- Can the learner use make-ten relationships at a larger scale?
- Can the learner recognise doubles and near doubles?
- Can the learner compensate in addition without changing the total?
- Can the learner compensate in subtraction with the correct correction direction?
- Can the learner find a small difference by counting up?
- Can the learner choose among strategies without a prompt?
- Can the learner explain why the strategy works?
- Can the learner estimate the answer’s size?
- Can the learner decide when a written algorithm is safer?
The deeper lesson: mental calculation is controlled transformation
Consider what the learner is actually doing in a strong mental calculation.
49 + 36 becomes 50 + 35.
63 − 29 becomes 63 − 30 + 1.
27 + 28 becomes 27 + 27 + 1.
61 − 58 becomes the gap from 58 to 61.
In each case, the original problem is transformed into an equivalent form whose useful structure is easier to see.
That is not a small arithmetic trick.
It is one of the central habits of mathematics.
Later students will factor algebraic expressions, substitute variables, transform graphs, rename fractions, choose coordinate systems and reformulate probability problems.
The scale becomes more sophisticated.
The habit is recognisable:
Preserve what matters. Change the form until the structure becomes easier to use.
Where this leads next
Flexible two-digit calculation prepares learners for three-digit mental arithmetic, multiplication through distributive reasoning, division through decomposition, estimation and multi-step word problems.
It also prepares an important examination habit:
do not automatically execute the first method you remember.
Look at the numbers first.
Sometimes the structure gives you a much shorter route.
For the wider mathematics map, see eduKateSG’s How Mathematics Works resources.
Final thought
A child who can calculate 49 + 36 vertically knows something useful.
A child who sees 50 + 35 knows something else as well.
The learner has noticed that the problem has more than one form.
One form is awkward.
Another makes the relationship almost obvious.
Choosing between them is the beginning of mathematical judgement.
Mental calculation is not about keeping everything in your head. It is about seeing enough structure that the calculation becomes worth keeping there.