PSLE Mathematics sometimes offers more than one valid route. A bar model, equation, unitary method, table, logical argument or backward method may all reach the same target. The best examination method is not always the shortest-looking one. A strong method is one the learner can execute accurately and verify.
This Learner’s Guide develops one advanced decision rule: choose the method you can verify. Before committing to a long route, ask how you will know the method is still valid halfway through and how you will test the final answer.
CHOOSE A METHOD → NAME WHAT IT PRESERVES → IDENTIFY THE CHECK → SOLVE → VERIFY WITH A DIFFERENT SIGNAL.
The three tests of a useful method
- Meaning: can you explain what each important number, bar, symbol or step represents?
- Progress: does each step move toward the target without creating unnecessary unknowns?
- Verification: can you test the result by substitution, inverse operation, estimation, units, a small case or a second representation?
Plan the verification before the final line
Learners often wait until the end before thinking about checking. A stronger habit is to choose the verification route when choosing the method. If you solve by algebra, plan to substitute. If you solve by percentage, plan to reconstruct the original relationship. If you solve by pattern, plan a small-case test.
Method pairings
Bar model → Equation or unit check
Why: Use the model for structure and a simple equation or target check for verification.
Equation → Substitution
Why: Put the answer back into the original relationship.
Percentage calculation → Reconstruct 100%
Why: Check that applying the stated percentage change returns the given amount.
Ratio unit method → Total or difference check
Why: Recombine the parts and confirm the original total or difference.
Speed problem → Distance-time reconstruction
Why: Check that speed × time reproduces the distance for each stage.
Average problem → Total reconstruction
Why: Multiply average by count, then verify the combined total.
Geometry → Alternative decomposition
Why: Split the shape differently or compare with a bounding shape.
Pattern rule → Small cases
Why: Test the rule on the first few terms before extending it.
Work backwards → Forward replay
Why: After reconstructing the start, replay the operations forward to reach the known end.
Logical elimination → Direct condition check
Why: Test the surviving option against every stated condition.
When the shortest method is not the safest
A very compact solution can hide fragile leaps. If the learner cannot explain why a shortcut works or cannot reconstruct the relationship after an error, the shortcut may be expensive under pressure. Use a method that is short enough but still transparent.
When the familiar method is not the right method
A favourite method should not be forced onto every problem. A bar model is strong for parts and wholes but may be awkward for a pattern. Algebra can be compact but fragile if the meaning of x drifts. Match method to structure and personal control.
Twelve method-choice cases
Ratio total
Situation: Bar model and unitary method both work.
Choice rule: Choose the route where one-unit meaning and total recombination are easiest to verify.
Percentage reverse
Situation: Direct algebra and 100-unit model both work.
Choice rule: Use the method that makes the base quantity clear and can reconstruct the final percentage.
Speed stages
Situation: Table and equations both work.
Choice rule: Choose the route that keeps each stage’s distance and time labelled and allows total reconstruction.
Average groups
Situation: Formula manipulation and total reconstruction both work.
Choice rule: Prefer the route that makes group totals visible if group sizes differ.
Geometry shaded area
Situation: Subtract-from-whole and add-components both work.
Choice rule: Choose the decomposition with fewer fragile measurements and an easy bounding check.
Pattern
Situation: Visual reasoning and formula rule both work.
Choice rule: Test either route on small terms before extending.
Fractions of remainder
Situation: Bar model and sequential equations both work.
Choice rule: Choose the method that keeps the changing whole visible.
Money rate
Situation: Unitary method and direct multiplication both work.
Choice rule: Use the route that preserves price-per-unit meaning and unit consistency.
Unknown start state
Situation: Forward trial and backward reconstruction both work.
Choice rule: Backward method may be easier because the final state is known; verify by replaying forward.
MCQ elimination
Situation: Full calculation and constraint elimination both work.
Choice rule: If constraints remove options safely, verify the survivor against the original conditions.
Repeated change
Situation: Percentage multiplier and staged amounts both work.
Choice rule: Use staged amounts if multiplier notation is fragile; verify by reconstructing each stage.
Composite quantity
Situation: One long expression or several named intermediates.
Choice rule: Prefer named intermediates if the long expression is hard to audit.
The method-verification card
- What is the target?
- What relationship must remain true?
- Which method makes that relationship visible?
- Where is the fragile step?
- How will I verify the result independently?
Do not verify with the same error
If the first method used the wrong copied value, repeating the same calculation will agree with itself. If the first method used the wrong percentage base, a second line using the same base is not independent. Verification should attack a different possible failure.
Verification under time pressure
Not every problem deserves a full second solution. Use the smallest independent check that can expose the important risk: units, magnitude, substitution, recombination, small case, or target identity.
Foundation recap: test the method on a small case before you commit
PSLE Mathematics questions sometimes look difficult because the numbers are large, the context is unfamiliar or several steps are mixed together. A learner may commit to a method too quickly, perform a long calculation and only discover at the end that the method never matched the relationship. One powerful way to test a method before investing heavily in it is to use a small case.
A small case keeps the structure of the problem but replaces difficult values with simple ones. It is not used to guess the final answer. It is used to ask: If my method were correct, would it behave sensibly in an easier version? If the method fails on 2, 3, 5 or 10 units, it is unlikely to become correct merely because the original numbers are larger.
This Learner’s Guide builds on Vol 0003: Represent Before You Calculate, Vol 0007: Recover When the First Method Fails, and Vol 0011: Name the Intermediate Quantity Before You Use It.
KEEP THE RELATIONSHIP → SHRINK THE NUMBERS → TEST THE METHOD → EXPLAIN WHY IT WORKS → RETURN TO THE ORIGINAL.
The quick answer: what is a small-case test?
A small-case test creates an easier version of the same mathematical structure. If a problem uses a ratio of 12:18, the learner might test the relationship with 2:3. If a pattern involves the 50th stage, test stages 1, 2 and 3. If a percentage relationship feels confusing, test it with a base of 100. If a repeated operation seems suspicious, try it with a tiny quantity where the result can be checked mentally.
The small case is a diagnostic tool. It tells the learner whether the proposed method preserves the relationship. After the method is understood, return to the actual numbers.
What must stay the same?
- Relationship: ratio, fraction, percentage, rate, difference, pattern or geometry relationship.
- Order of events: if the original has increase then decrease, the small case must preserve that order.
- Target type: if the original asks for a total, the small case should also test a total.
- Conditions: restrictions such as whole-number counts or equal groups must remain meaningful.
- Units in principle: the small case can simplify values, but it should not turn an area problem into a length problem.
If the learner changes the structure while simplifying the numbers, the test no longer diagnoses the original method.
When a small case is useful
- The method feels abstract and the learner cannot explain why it works.
- A pattern rule is being proposed from several stages.
- A ratio or fraction relationship is hidden by large numbers.
- A repeated percentage or change process is confusing.
- A method produces an impossible result and the learner wants to locate the structural error.
- Two methods seem plausible and the learner wants to compare them quickly.
When not to use it
Do not create a small case when the original numbers themselves are essential to the relationship, when a direct method is already clear and efficient, or when simplifying destroys an important condition. A small case should clarify structure, not create extra work.
Do not turn every PSLE Mathematics question into an experiment. The skill is most useful for difficult or uncertain methods.
Ratio: shrink the units, not the meaning
Suppose two quantities are in the ratio 12:18. The common structure is 2:3. Testing with 2 red counters and 3 blue counters makes the part-to-part relationship visible. If a proposed method says the total is 6 units instead of 5, the error becomes obvious before large-number arithmetic begins.
Worked ratio test
Original idea: A:B = 4:7 and the total changes after 3 units are added to A. Before using large values, test with A=4 and B=7. Add 3 to A and inspect the new ratio. The learner sees immediately that changing one part does not preserve the original ratio. The small case exposes a common false assumption.
Fractions: use a whole that divides cleanly
If a problem says one-third of a quantity is removed and then one-quarter of the remainder is used, choose a small whole such as 12. Remove one-third: 4 removed, 8 remain. Then use one-quarter of the remainder: 2. This shows that the second fraction applies to 8, not to the original 12.
The small case is especially helpful when the whole changes between steps.
Percentage: use 100 when the base is confusing
Percentage relationships often become clearer with a base of 100. If a quantity increases by 20% and then decreases by 20%, test 100 → 120 → 96. The result shows that the changes do not cancel because the two percentages use different bases.
Why this matters
A learner who merely subtracts 20% from 20% may be using symbolic appearance instead of relationship. The small case turns the relationship into visible quantities.
Average: use tiny groups
If a learner wants to average two averages directly, create a small counterexample. Group A: one score of 10, average 10. Group B: nine scores averaging 20. The combined average is not 15 because the groups are unequal in size. The small case exposes the method failure.
Speed: use easy times
If a journey has two different speeds, test with one hour at 10 km/h and one hour at 20 km/h. Total distance is 30 km over 2 hours, so average speed is 15 km/h. Then change the times: one hour at 10 and three hours at 20. The simple average of speeds is still 15, but the true average is different. This reveals why time weighting matters.
Patterns: inspect stages 1, 2 and 3
Before applying a formula to stage 50, test the rule on the first few stages. If the formula fails stage 2, it should not be trusted for stage 50. Small cases are one of the strongest ways to test pattern rules.
Worked pattern case
A learner thinks the nth figure contains 3n tiles. Stage 1 has 4 tiles, stage 2 has 7 and stage 3 has 10. The proposed rule gives 3, 6 and 9, so it fails immediately. The learner can then look for the correct structure, such as 3n+1.
Geometry: use a simpler shape arrangement
If a composite-area method feels unclear, test it on a rectangle split into two smaller rectangles with easy dimensions. Does adding component areas recover the whole? Does subtracting the cut-out produce the shaded region? The simple geometry verifies the operation logic before the learner handles harder dimensions.
Perimeter: small cases expose double-counting
When two rectangles are joined, learners sometimes add both full perimeters and forget that the shared edge becomes internal. Use two 1-by-2 rectangles. Draw them joined. The shared edge is visibly not part of the outside perimeter. The small case reveals what must be subtracted or avoided.
Volume: test one layer first
For stacked cubes or repeated layers, calculate one simple layer and then two layers. If the proposed method doubles the wrong dimension or counts shared structure incorrectly, the small model exposes it.
Combinations and counting: list tiny cases
If a counting method is uncertain, use a case with two or three choices and list every valid outcome. Compare the list with the formula or multiplication rule. If they disagree, inspect the logic before scaling up.
Whole-number constraints
Some contexts require whole-number answers: people, books, buses, teams. A small case can reveal whether a method produces impossible fractions. If a proposed sharing rule gives 2.5 pupils in a group, the method or interpretation needs review.
The small-case method is not guessing
The learner must be able to explain why the small case preserves the original structure. Simply trying random numbers until an answer looks nice is not mathematical reasoning.
A valid small case is chosen deliberately to simplify arithmetic while keeping the relationship.
Thirty small-case tests
Ratio total
Small case: Test 2:3 before 24:36.
What it tests: Verify that total units are 5, not the numerical difference or product.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Ratio change
Small case: Use 2:3, then add 1 unit to the first part.
What it tests: See that the original ratio changes when only one part changes.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Fraction of remainder
Small case: Use 12 for one-third removed then one-quarter of remainder.
What it tests: Confirm the second fraction uses the new whole.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Repeated percentage
Small case: Use 100 for +10% then -10%.
What it tests: Show that equal percentages on different bases do not cancel.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Discount then tax
Small case: Use $100 before discount and later increase.
What it tests: Track the changing base explicitly.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Average unequal groups
Small case: Use 1 item and 3 items.
What it tests: Show why averaging averages can fail.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Average equal groups
Small case: Use two groups with equal counts.
What it tests: Show when averaging the group averages does work.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Speed equal times
Small case: Use 1 hour at each speed.
What it tests: Compare simple average with total-distance/total-time method.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Speed unequal times
Small case: Use 1 hour and 3 hours.
What it tests: Expose why simple averaging fails.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Pattern formula
Small case: Test stages 1, 2 and 3.
What it tests: Reject any formula that fails early stages.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Growing difference
Small case: Use a sequence with differences 2, 4, 6.
What it tests: Check whether a proposed constant-difference rule is false.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Perimeter joined shapes
Small case: Join two 1-by-2 rectangles.
What it tests: See that shared edges are internal.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Area subtraction
Small case: Use a 4×4 square with a 1×1 cut-out.
What it tests: Verify whole-minus-cut-out logic.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Volume layers
Small case: Use one 2×2 layer, then two layers.
What it tests: Check whether multiplying by number of layers is valid.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Unit conversion
Small case: Use 1 m = 100 cm before large lengths.
What it tests: Test direction of multiplication or division.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Rate per item
Small case: Use $2 per book for 3 books.
What it tests: Check whether total = rate × count.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Rate reverse
Small case: Use $6 total for 3 books.
What it tests: Check whether unit price = total ÷ count.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Sharing
Small case: Share 12 items among 3 groups.
What it tests: Verify division meaning before large values.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Grouping
Small case: Make groups of 3 from 12 items.
What it tests: Compare grouping interpretation with sharing.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Difference model
Small case: Use 5 and 8.
What it tests: Check whether total, difference and part are being confused.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Before-and-after
Small case: Start with 10, add 4, remove 3.
What it tests: Track stage order before larger values.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Algebra-like unknown
Small case: Use x+3=8 with small values.
What it tests: Check whether inverse operation is applied in correct direction.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Table pattern
Small case: Use two rows before ten rows.
What it tests: Test whether the row relationship remains consistent.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Discrete count
Small case: Use 5 pupils.
What it tests: Reject methods that create half a pupil when whole counts are required.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Probability intuition
Small case: Use two coins or a tiny equally likely set.
What it tests: List outcomes instead of relying on impression.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Fraction comparison
Small case: Use halves and quarters of the same small whole.
What it tests: Check whether numerator-only comparison fails.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Scale drawing idea
Small case: Use a 1:2 scale with a 3 cm object.
What it tests: Verify direction of scaling.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Time intervals
Small case: Use 10:00 to 10:30.
What it tests: Check elapsed-time method before crossing hours.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Money change
Small case: Use $10 purchase from $20.
What it tests: Verify subtraction direction and meaning.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
Multi-step target
Small case: Use tiny values and label each intermediate.
What it tests: Confirm the final step answers the target rather than stopping early.
After the method passes the small case, return to the original numbers and explain why the same relationship still applies. A small case validates structure; it does not replace the original calculation.
The small-case protocol
- State the original relationship in words.
- Choose the smallest numbers that preserve it.
- Predict what a correct method should do.
- Run the proposed method.
- Compare the result with the prediction or a directly checkable outcome.
- If it passes, return to the original; if it fails, repair the method.
A failed small case is useful information
Do not feel frustrated when the small case disproves a method. That is the purpose of the test. It has prevented a longer wrong solution. Use Vol 0007: return to the representation and find the broken relationship.
A passing small case is not a full proof
One successful small case increases confidence but does not prove every possible method is correct in every situation. The learner should still explain why the method preserves the original relationship. A rule that works for one special case may fail elsewhere.
For patterns, test more than one early stage. For methods involving different conditions, choose cases that expose the critical relationship.
Small cases and answer checking
A small case can also act as an independent check. If a formula or general method was derived algebraically, test it with one simple value. This is different from repeating the original arithmetic and can expose structural mistakes.
Small cases under time pressure
Use them selectively. A ten-second mental case can be enough. If the method is already obvious and stable, do not create extra work. If the method is uncertain and the calculation would be long, a small case can save time.
The one-minute method test
For a difficult problem, give yourself one minute in practice to build a tiny version, test the method and state the conclusion: “works because…”, “fails because…”, or “unclear because…”. This trains fast structural checking.
How this connects to intermediate quantities
Name each small-case result just as you would in the original problem. If 5 is the total number of ratio units, say so. If 20 is the discounted price, label it. Vol 0011’s quantity-control habit prevents the small case from becoming anonymous arithmetic.
A seven-day small-case cycle
- Day 1: ratio and fraction cases.
- Day 2: percentage and changing-base cases.
- Day 3: average and speed cases.
- Day 4: patterns and tables.
- Day 5: geometry, perimeter and volume.
- Day 6: mixed word problems under gentle timing.
- Day 7: delayed transfer to unfamiliar contexts.
What parents and tutors should ask
Ask: What did you keep the same? What did you simplify? Why is this still the same mathematical relationship? What result should a correct method produce in the small case? What did the test reveal?
These questions prevent the exercise from turning into random-number play.
Common mistakes
- Changing the structure: simplifying away the relationship being tested.
- Guessing with random numbers: trying values without a reason.
- Using one special case as proof: a passing example is evidence, not universal proof.
- Forgetting the original: solving the small case correctly but never returning to the real numbers.
- Testing arithmetic instead of method: the small case should reveal structure, not just make multiplication easier.
- Overusing the tool: creating small cases for routine questions that already have a clear route.
Frequently asked questions
Can I use a small case in the actual examination?
Yes as scratch reasoning when useful, but the final solution must still answer the original question clearly.
What numbers should I choose?
Choose values that make the relationship easy to inspect: 1, 2, 3, 5, 10, 12 or 100 are common choices, depending on the structure.
What if the small case works but the original answer is still wrong?
The method may be structurally valid but the original execution, units, copying or target may be wrong. Diagnose the next layer.
Is this the same as estimation?
No. Estimation checks approximate magnitude. A small case checks whether a method or relationship behaves correctly on simpler values.
Is this the same as using examples to prove a rule?
No. Examples can test and challenge a proposed rule, but one or a few examples do not necessarily prove a general statement.
Foundation recap: represent before you calculate
PSLE Mathematics performance becomes more reliable when the learner can see the problem before calculating it. Many errors begin with an operation chosen too early: multiply because there is a percentage, divide because there is a total, subtract because something decreased. The safer habit is represent before you calculate.
This guide develops the Mathematics branch of Vol 0001: Read Before You Solve. For the larger topic map, use the Primary 6 Mathematics Learning Hub and the PSLE Learning Guide.
READ → NAME THE QUANTITIES → SHOW THE RELATIONSHIP → CHOOSE A METHOD → COMPUTE → CHECK THE RESULT.
Why representation comes before calculation
A mathematical question may contain perfectly familiar numbers inside an unfamiliar relationship. If the learner begins calculating before identifying that relationship, correct arithmetic can produce the wrong answer.
Representation means making the structure visible. It can be a bar model, diagram, table, equation, ratio statement, number line, unit-rate statement, annotated figure, list of cases or simply a carefully written sentence describing what is known and unknown.
The representation should reduce confusion. It is not an extra decoration.
The three questions to ask before touching the calculator or doing arithmetic
- What are the quantities?
- How are they related?
- Which quantity am I actually asked to find?
These three questions stop many common errors because they separate the mathematical situation from the operations used to solve it.
Example 1: percentage — increase by is not increase to
Suppose a quantity is 240 and increases by 25%. A rushed learner may write 240 × 25% = 60 and stop. The calculation 60 is correct, but it is the increase, not the new total. The representation should make the relationship explicit: original 100% → increase 25% → new total 125%.
Now the learner can decide whether the question asks for the amount of increase or the final quantity. The mathematics becomes a task-selection problem before it becomes arithmetic.
Example 2: ratio — the numbers are labels for a relationship
If the ratio of red to blue beads is 3:5 and there are 40 blue beads, the number 5 corresponds to 40. One part is 8. Red is 3 parts, so red is 24. The useful representation is not merely “3:5”. It is 5 parts = 40 → 1 part = 8 → 3 parts = 24.
When the relationship is visible, the operation sequence has a reason.
Example 3: average — protect the total
Average questions are often easier when the learner converts average into total. If the average of six values is 18, the total is 108. A changed average after adding, removing or replacing a value should be reasoned through totals, not by manipulating averages as if they were independent quantities.
AVERAGE × NUMBER OF ITEMS = TOTAL.
This representation turns a vague average problem into conservation of total quantity.
Example 4: speed — label the unit relationship
Speed is a rate: distance per unit time. Before choosing a formula, name the three quantities and their units. If a journey has two stages with different speeds, the overall average speed is not generally the simple average of the two speeds. Represent each stage through distance and time, then combine totals.
The correct formula matters, but the representation explains when it applies.
Choose the simplest useful representation
- Use a bar model for part-whole, comparison, ratio, before-after and many fraction/percentage relationships.
- Use a table when several cases, categories or paired values must stay aligned.
- Use an equation when an unknown quantity has a clear algebraic relationship.
- Use a diagram for geometry, movement, spatial arrangements or overlapping regions.
- Use a number line for ordered values, differences, intervals and some fraction/decimal reasoning.
- Use systematic listing when all valid cases must be counted without omission or duplication.
- Use a unit-rate statement when a “per one” relationship controls the problem.
Do not force a favourite method onto every problem. Representation is successful when it makes the controlling relationship clearer.
The “operation reflex” trap
Learners often memorise cue words: “altogether means add”, “difference means subtract”, “of means multiply”. These can help at very basic stages, but they are unreliable in complex problems because the same word can appear in different structures.
Replace cue-word guessing with relationship reading. Ask what is being combined, compared, scaled, shared, repeated or changed.
A complete PSLE Mathematics launch
- Read the final question and identify the target quantity.
- List or mark the given quantities with units.
- State the important relationship in words.
- Draw or write the smallest useful representation.
- Estimate the rough size or direction of the answer if possible.
- Choose the method.
- Compute carefully.
- Attach the correct unit and answer the stated question.
- Check using estimation, inverse operation, substitution or a second representation when appropriate.
Why estimating before solving is powerful
An estimate creates a boundary. If the exact answer later falls far outside that boundary, the learner has evidence that something went wrong. This catches calculator slips, place-value mistakes, reversed ratios and impossible measurements.
Estimation does not need to be precise. It needs to be informative.
Checking without redoing the whole problem
1. Unit check
Does the answer have the unit the question requires? If the question asks for area and the answer is in centimetres instead of square centimetres, the final line is already unstable.
2. Magnitude check
Is the answer sensible compared with the starting quantities? A discount should not usually make the final price larger. A part should not exceed a total unless the context allows it.
3. Inverse check
If you divided to find one part, multiply back. If you solved an equation, substitute the value. If you found a percentage of a whole, compare it with the whole.
4. Structural check
Return to the model or relationship. Did you answer the target quantity or an intermediate quantity?
When a difficult word problem feels blank
Do not immediately search memory for a matching worksheet. Break the question into stable information.
- What is fixed?
- What changes?
- What is being compared?
- What is before and what is after?
- What is equal, proportional or conserved?
- Can one unknown be expressed in terms of another?
These questions often reveal a structure even when the surface story is unfamiliar.
A worked mixed problem routine
Imagine a container is partly filled. Some liquid is removed, then water is added, and the final mixture has a stated fraction of one component. The story has several events, so do not calculate from the first sentence. Define the original total, track what is removed, track what remains, then represent the final mixture. The problem becomes a before-after conservation structure.
The important habit is not the specific method. It is refusing to let chronology hide the quantities.
The five Mathematics error families
- Representation error: the relationship was modelled incorrectly.
- Strategy error: the representation was reasonable but the chosen method could not reach the target.
- Execution error: arithmetic, algebra or calculator work was inaccurate.
- Communication error: working, labels, units or final answer were unclear or incomplete.
- Checking error: an impossible or wrong-target answer survived because it was never tested.
The repair depends on the family. More practice questions do not automatically repair a representation error.
Practice progression: basic to advanced
- Basic: identify target quantity and units before solving routine questions.
- Foundation: draw or write the relationship for ratio, percentage, fraction and average questions.
- Core: solve mixed questions where the topic is not labelled.
- Transfer: solve changed-context questions with the same underlying relationship.
- Advanced: compare two valid methods and explain why each works.
- Exam control: decide when to move on, when to check, and when a representation needs to be rebuilt rather than patched.
When the skill is becoming independent
- The learner can explain what each number represents before using it.
- The learner can choose between a bar model, table, equation or diagram rather than drawing automatically.
- The learner notices when an intermediate result is not the final answer.
- The learner estimates and catches unreasonable results.
- The learner can solve the same relationship in a changed context.
- The learner can recover from a failed method by returning to the representation instead of guessing another operation.
Next route
Continue to Vol 0004: Science — Evidence Before Explanation, return to Vol 0001 for the shared PSLE launch routine, or use the Primary 6 Mathematics Learning Hub for the wider Mathematics branch.
Official examination reference
For the current assessment objectives and format, use the correct examination-year document from the Singapore Examinations and Assessment Board. For 2026, see PSLE Mathematics. The official document and school instructions take priority over generic study advice.
Independence indicators
- The learner can create a small case without changing the structure.
- Long wrong methods are caught earlier.
- Pattern rules are tested on early stages.
- Changing-base errors become easier to detect.
- The learner can explain why a passing small case supports the method.
- The learner returns to the original problem instead of stopping at the test case.
Next route
Continue to Vol 0016: Science — Use a Counterexample to Challenge an Always-or-Never Claim.
Official PSLE reference
SEAB’s PSLE page and PSLE Formats Examined in 2026 are the current official examination references. Official examination documents and school instructions take priority over generic study advice.
Series: How to Perform in PSLE | Learner’s Guide · Vol 0015 · Intermediate Mathematics method validation
Next route
Continue to Vol 0025: Science — Compare the Change, Not Just the Final Value.
Official PSLE reference
For current examination-year information, use the official SEAB PSLE pages and the relevant current-year formats. Official examination documents and school instructions take priority over generic study advice.
Series: How to Perform in PSLE | Learner’s Guide · Vol 0024 · Advanced Mathematics method control
