Posting Group 3 can make Mathematics look straightforward: the student enters secondary school through PG3, studies Mathematics at G3, and heads toward the SEC examination. But the label is not the mechanism. Mathematics works through a system of concepts, representations, procedures, reasoning, modelling, communication and problem solving that must become increasingly independent under unfamiliar conditions.
Alicia may execute algebra accurately but fail to translate a word problem into an equation. Tricia may understand the concept but choose a method that is too slow under examination time. Kai Kai may obtain the correct numerical answer yet lose marks because the reasoning, units, interpretation or mathematical argument is incomplete. All three can be “good at Mathematics” and still have very different readiness profiles.
For 2027 school candidates, SEAB lists G3 Mathematics as syllabus K310. The current syllabus is organised around Number and Algebra, Geometry and Measurement, and Statistics and Probability, while also emphasising mathematical processes such as reasoning, communication, application and modelling. Its assessment objectives distinguish standard techniques, problem solving in varied contexts, and mathematical reasoning/communication. That destination map matters because it shows that G3 Mathematics is not merely faster calculation.
This article owns the PG3 Mathematics mechanism job. It does not replace the broader How Mathematics Works owner, the existing G1/G2/G3 Mathematics pathway pages or Secondary 4 SEC guides. Instead, it explains how a PG3 learner should build the mathematical engine that can survive increasing G3 demand.
Posting Group 3 is an entry route under Full Subject-Based Banding. It is not a permanent statement of mathematical capability. Subject-level arrangements follow current school and MOE processes. This guide concerns learning and readiness, not an administrative guarantee.
Alicia, Tricia and Kai Kai are fictional learners. Their problems, marks and worked examples are original teaching material rather than official SEAB questions or school placement tests.
01. PG3 is an entry route, not a mathematical identity
Posting Group 3 describes entry into secondary school under Full Subject-Based Banding. It does not mean every student enters with identical mathematical fluency, and it does not turn G3 Mathematics into a personality trait.
Two PG3 learners can differ sharply. One may have excellent calculation but fragile modelling. Another may understand concepts deeply but work too slowly. A third may reason well yet make frequent notation or sign errors.
Use a mathematical profile
Instead of “PG3 Math student”, describe current capability: “algebraic manipulation secure; graph interpretation strong; equation formulation inconsistent; reasoning explanations brief; timed mixed-paper completion improving.”
Keep posting group and subject level separate
Full SBB allows a mix of subject levels under applicable arrangements. The instructional question is therefore what the learner can do mathematically now, not what the posting label predicts forever.
Marks need mechanism
A 75% mark can hide incomplete reasoning, careless checking or one severe modelling weakness. A 60% mark can contain strong reasoning but weak speed. The next lesson should be chosen from the error pattern beneath the percentage.
Mathematics is cumulative but not linear
Later topics reuse earlier ideas in new combinations. Fractions reappear in algebra. Ratio appears in similarity and probability. Graphs connect to equations, rates and modelling. Weak links can therefore emerge far from the chapter in which they were first taught.
Formal pathway decisions remain external
A tutor can document mathematical readiness and independence. The school applies current progression arrangements. This article builds capability rather than inventing private placement rules.
02. The mathematical engine: concept, representation, method, reasoning
A strong mathematical solution contains more than a procedure. It begins with a concept, represents the problem usefully, selects a method, executes accurately, interprets the result and checks whether the answer makes sense.
Concept
What relationship is operating? Proportion? Linear change? Area? Probability? Algebraic equivalence? Without the concept, the learner searches for surface clues.
Representation
Can the situation be expressed as an equation, graph, table, diagram, ratio or coordinate model? Good representation often reduces difficulty before calculation begins.
Method
Several methods may work. The best method depends on efficiency, reliability and the information given.
Execution
Algebraic manipulation, arithmetic, substitution, construction and calculation need procedural fluency. Reasoning cannot compensate for repeatedly incorrect arithmetic.
Interpretation
A numerical result must be returned to the context. If the model gives 3.6 buses, the practical answer may require 4 buses. If a length becomes negative, the algebra may be valid but the physical interpretation impossible.
Checking
A check should use a different viewpoint where possible: substitute back, estimate scale, compare units, use an alternative method, or inspect a graph.
Reasoning
Why is the method valid? Why does the pattern hold? Why can a statement be concluded? Mathematical communication makes those relationships inspectable.
Diagnosis follows the engine
If Alicia knows the concept but chooses the wrong representation, the repair differs from Tricia, who chooses well but executes poorly. Kai Kai may execute and interpret correctly yet fail to justify.
03. How 2027 G3 Mathematics K310 should be read
SEAB lists Mathematics K310 for 2027 G3 school candidates. The syllabus organises content into Number and Algebra, Geometry and Measurement, and Statistics and Probability, while emphasising reasoning, communication, application and modelling.
AO1 · Standard techniques
The learner must recall and use facts, terminology and notation, read direct information from tables, graphs, diagrams and texts, and carry out routine procedures accurately.
AO2 · Solve problems in varied contexts
The learner must identify relevant mathematics, translate between forms, connect topics, formulate problems mathematically, select information and techniques, and interpret results in context.
AO3 · Reason and communicate mathematically
The learner must justify statements, explain results in context and write mathematical arguments.
The official syllabus gives approximate assessment weightings of 45% for AO1, 40% for AO2 and 15% for AO3. The implication is important: routine technique remains essential, but a large share of performance depends on choosing and using mathematics in context and communicating reasoning.
Do not project the examination backwards mechanically
Lower-secondary teaching should build the concepts and processes that later support K310. The examination destination should guide readiness without reducing every Secondary 1 or Secondary 2 lesson to paper drilling.
Read strands as connected
Algebra can model geometry. Graphs can represent statistics. Ratio appears inside trigonometry. Probability depends on fraction sense and systematic counting. The real system is interconnected.
04. Problem solving is selection, not just execution
A student can know many procedures and still struggle with G3 Mathematics because unfamiliar problems do not announce which procedure to use.
Routine questions name the road
“Solve the quadratic equation” tells the learner what object is present. A contextual problem may require the learner to recognise that a quadratic relation exists before any solving begins.
Selection is a mathematical act
Choosing between simultaneous equations, ratio, gradient, Pythagoras or probability requires the learner to interpret structure.
Surface words can mislead
The word “increase” does not automatically mean percentage. “Speed” does not automatically mean distance divided by time if the question concerns a graph or average rate over intervals. The relationship determines the method.
Use a first-minute routine
What is given? What is required? Which quantities are related? Which representation makes the relationship visible? Which method follows from that representation?
Train method comparison
Solve one problem two ways and compare reliability. Graphical and algebraic methods, elimination and substitution, exact and approximate reasoning can each be useful in different contexts.
Do not overtrain pattern matching
If every worksheet groups identical question types, the learner can succeed through recognition of layout rather than mathematical selection. Mixed practice is necessary once the method is learned.
Problem solving includes interpretation
A correct equation can produce an answer that is impossible in context. The final step belongs to Mathematics too.
05. Diagnose the first weak mathematical mechanism
A wrong answer is not a diagnosis. The first incorrect mathematical decision is the useful target.
Example · The rectangular garden
A rectangular garden has area 96 m². Its length is 4 m longer than its width. Find its dimensions.
Alicia writes 96 ÷ 4 = 24 and stops. First failure: representation. She has not converted the relationship between length and width into a model.
Tricia lets width = x and length = x + 4, forms x(x+4)=96 correctly, but expands incorrectly. First failure: algebraic execution.
Kai Kai solves x²+4x−96=0 and obtains x=8 or x=−12, but reports both as possible widths. First failure: contextual interpretation.
Use five diagnostic questions
What concept applies? What representation was chosen? Was the method valid? Was execution accurate? Was the answer interpreted and checked?
Record support
Did the learner need the variable defined for them? Did the tutor suggest using an equation? Did the calculator step need correction? Support changes the meaning of a correct final answer.
Retest with changed surface details
After repair, use a rectangle with perimeter relation, a consecutive-number problem or another formulation task. The learner should reconstruct the modelling step.
06. Number sense, accuracy and estimation
G3 Mathematics still depends on number sense. Algebra, geometry and statistics all collapse if the learner cannot judge scale, sign, magnitude and reasonableness.
Accuracy is necessary but not sufficient
A calculator can produce a decimal precisely from incorrect input. Mathematical control begins before the keypress: estimate the order of magnitude, check sign and know what quantity is being calculated.
Fractions remain structural
Fractions are not a Primary-school relic. They appear in algebraic expressions, probability, rates, gradients and ratios. A learner who avoids fraction manipulation may compensate with decimals until exact work becomes inconvenient.
Percentage is multiplicative
Successive percentage changes do not cancel symmetrically. A 20% increase followed by a 20% decrease gives 1.2 × 0.8 = 0.96 of the original amount, a 4% net decrease.
Use estimation before exact calculation
If 19.8 × 5.1 is entered, an estimate near 20 × 5 = 100 gives a reasonableness check. A displayed answer of 1000 should trigger review immediately.
Track units
Rate, area, volume and percentage problems often reveal misunderstanding through units. A numerical answer with the wrong unit can expose a deeper representation error.
Negative quantities need context
Negative numbers can represent direction, change, debt or algebraic roots. Whether a negative result is meaningful depends on the context.
Bounds and approximation need intention
Rounding too early can distort a later result. Keep suitable precision through the working, then report appropriately at the end.
Number sense supports checking
Before trusting an answer, ask whether it is too large, too small, wrong in sign or impossible relative to the data.
Recovery evidence
A stronger learner catches impossible outputs before the teacher does and uses estimation naturally rather than only when a question explicitly asks for it.
07. Algebra as structure, not symbol pushing
Algebra is a language for relationships. A learner who sees only symbols may memorise procedures without understanding why expressions remain equivalent.
Variables can represent quantities or general relationships
In one problem x is an unknown width. In another it varies along a graph. In an identity it can represent any permitted value. The role matters.
Equality means balance, not “the answer comes next”
2x + 3 = 11 states that two expressions have the same value. Solving preserves that equality through valid operations.
Manipulation should preserve equivalence
Factorisation, expansion, simplification and rearrangement are not visual tricks. They produce equivalent forms useful for different jobs.
Choose form by purpose
Expanded form may reveal coefficients; factorised form may reveal roots; completed-square form may reveal a turning point. Algebraic fluency includes choosing a useful representation.
Watch invisible structure
In 3(x−2), the bracket matters. In (x+2)/(x−3), restrictions matter. Many “careless” errors are actually failures to see grouping.
Use substitution as a check
If two expressions are claimed equivalent, test a simple permitted value during practice. This does not prove identity universally, but it can expose a bad transformation quickly.
Algebra connects topics
Geometry produces equations from lengths and areas. Statistics uses algebraic relationships in formulas. Graphs translate algebra into shape.
Recovery evidence
A stronger learner can explain why a transformation is valid and select algebraic form based on the next mathematical job.
08. Equations, inequalities and formulation
Solving equations is only half the skill. G3 problem solving often requires the learner to formulate the equation from a situation.
Define variables clearly
“Let x be the width in metres” reduces later ambiguity. In simultaneous equations, define both quantities before forming relationships.
Translate sentence structure into mathematical structure
“The larger number is 7 more than twice the smaller” becomes L = 2S + 7, not L = 2(S + 7).
Use simultaneous equations when two unknown relationships interact
If adult and student tickets total 34 and revenue gives another equation, the pair of relationships can determine both counts.
Quadratics often emerge from products
Area, consecutive integers, projectile-like models and geometric relationships can produce quadratic equations. Recognising this structure matters before choosing factorisation or formula.
Inequalities describe ranges
“At least” and “no more than” are mathematical constraints. The solution is often an interval rather than one value.
Interpret roots
Algebra may give two roots. Context may allow both, one or neither. A negative time or impossible length must be rejected with reason.
Check by substitution
Substitute solutions into the original relationship, especially after fractional or quadratic manipulation.
Recovery evidence
A stronger learner can build equations from unfamiliar wording, solve them reliably and return the roots to the context rather than stopping at the algebra.
09. Functions and graphs as relationships
A graph is not a picture appended after algebra. It is another representation of a relationship between variables.
Read axes before shape
Identify quantity, unit and scale. A rising graph of “time taken” can mean a process is slowing, while a rising graph of “rate” means the opposite.
Gradient has meaning
For a linear relationship, gradient measures vertical change per horizontal change. In context, its units and meaning matter.
Intercept has meaning too
The y-intercept may represent a starting fee, initial value or model constant. Do not treat it as a decoration on the equation.
Quadratic graphs encode roots and turning points
Factorised form connects to x-intercepts. Completed-square form connects to maximum or minimum. Representation choice reduces cognitive load.
Exponential relationships differ from linear ones
Equal increments in x can produce multiplicative changes in y. Learners should compare structures rather than memorise curve shapes only.
Tangent gradient approximates instantaneous change
Where required, drawing a tangent to a curve estimates local gradient. The idea links graphical geometry with rate.
Use graphs to check algebra
Solutions of equations can appear as intersections or intercepts. A sketch can expose an impossible claimed number of roots.
Recovery evidence
A stronger learner moves between equation, table and graph without treating them as unrelated chapters.
10. Ratio, rate, percentage and proportional reasoning
Proportional reasoning is one of the most reusable structures in secondary Mathematics. It appears in scale, speed, density, similarity, percentages, financial contexts and probability.
Distinguish additive from multiplicative change
If every ticket costs $3 more, the change is additive. If every quantity increases by 15%, the change is multiplicative.
Unit rate reveals structure
Converting to cost per item, distance per unit time or mass per unit volume can make comparisons transparent.
Ratios compare relative quantities
A ratio 2:3 does not mean a difference of 1 in absolute terms. Scaling both parts preserves the ratio.
Direct proportion has a constant multiplicative relationship
If y ∝ x, then y/x is constant. A straight line through the origin can represent that relation under appropriate axes.
Inverse relationships require different intuition
If y varies inversely with x, doubling x halves y under the model. Learners should not force every proportional problem into direct proportion.
Percentage points are not percentage change
A rate moving from 40% to 50% rises by 10 percentage points but by 25% relative to the original 40%.
Recovery evidence
A stronger learner recognises when a problem is multiplicative, chooses a useful unit rate or ratio and interprets percentage language precisely.
11. Geometry and measurement
Geometry becomes powerful when learners stop treating diagrams as pictures to stare at and start reading them as constrained mathematical systems.
Diagrams encode relationships
Parallel lines, equal lengths, right angles, tangencies and symmetry are not decorative marks. Each can trigger a theorem or structural relationship.
Draw when the diagram is missing
Word problems involving bearings, loci, area or similarity become easier when the learner externalises the spatial information.
Area and volume require dimensional thinking
Length scales linearly, area quadratically and volume cubically. A scale factor of 3 produces an area factor of 9 and volume factor of 27 in similar figures.
Units reveal dimension
cm, cm² and cm³ represent different quantities. Unit conversion errors often reveal that the learner is manipulating numbers without tracking dimension.
Use theorem conditions carefully
Pythagoras applies to right-angled triangles. Similarity requires corresponding relationships. Circle results depend on specific configurations. Naming a theorem without verifying its conditions is not reasoning.
Measurement carries approximation
Real measurements can be rounded or bounded. A diagram drawn to scale may help estimation but should not replace exact reasoning unless the task allows measurement.
Coordinate geometry connects algebra and space
Gradient, distance and midpoint translate geometric relationships into algebraic form. The learner should move between the visual and symbolic versions.
Recovery evidence
A stronger learner marks relevant relationships, chooses the appropriate theorem and can explain why its conditions are satisfied.
12. Trigonometric and spatial reasoning
Trigonometry is often reduced to choosing a formula. The deeper skill is identifying which sides, angles and relationships define the geometry.
Orient the triangle first
Opposite and adjacent depend on the chosen angle. Label before substituting.
Use the relationship that matches the knowns and unknown
Sine, cosine and tangent are not three buttons to try randomly. Each compares specific side ratios in a right triangle.
Interpret inverse trigonometric output
The calculator can return an angle, but the learner must decide whether it is geometrically possible and whether degrees mode is appropriate.
Bearings require directional discipline
Three-figure bearings are measured clockwise from north. A clean sketch prevents many sign and orientation errors.
Three-dimensional problems need decomposition
A 3D shape can often be solved by identifying right triangles on faces or inside the solid. Spatial reasoning becomes manageable when the hidden two-dimensional structure is exposed.
Use exact relationships before decimal approximation
Keep symbolic or full calculator precision until the final answer where possible, especially across multi-step trigonometric work.
Recovery evidence
A stronger learner chooses the correct triangle and relationship before touching the calculator and can justify the geometry that makes the calculation valid.
13. Statistics and data interpretation
Statistics is not only calculation of mean, median and range. It is the interpretation of data in context, including what summaries reveal and what they hide.
Choose a summary suited to the data
The mean uses all values and is sensitive to extremes. The median is resistant to outliers. Range describes spread crudely. Different summaries answer different questions.
Read graphs critically
Scale, interval, truncation and category choice can shape visual impression. A graph can be numerically correct and still exaggerate or obscure differences.
Compare distributions, not single numbers
Two groups can share the same mean and have very different spread. When comparing performance, discuss both centre and variation where appropriate.
Understand sampling limits
A survey of one class does not automatically represent an entire school. Sample method matters to the strength of the conclusion.
Correlation requires caution
Two variables moving together does not establish causation. Other variables may explain the association.
Use context after calculation
If the mean waiting time is 7.4 minutes, explain what that means for the scenario rather than leaving the number isolated.
Recovery evidence
A stronger learner can calculate accurately and also explain what the summary, graph or sample does and does not support.
14. Probability and sample spaces
Probability rewards systematic representation. Many errors come from incomplete counting or assuming events are equally likely when they are not.
Build the sample space
Tables, tree diagrams and organised lists help expose all possible outcomes.
Distinguish independent and dependent events
With replacement, probabilities may remain unchanged. Without replacement, later probabilities can depend on earlier outcomes.
Use complements
“At least one” is often easier through 1 − P(none). The complement is a strategic representation, not a memorised trick.
Check probability bounds
Any probability below 0 or above 1 is impossible. This gives an immediate error signal.
Experimental and theoretical probability differ
Observed frequency can approach theoretical expectation over many trials but need not match exactly in a small sample.
Conditional language matters
“Given that”, “if” and “after” can change the sample space. Read the condition before calculating.
Recovery evidence
A stronger learner represents outcomes systematically and can explain why the chosen denominator matches the current sample space.
15. Mathematical modelling and assumptions
Mathematical modelling turns a real situation into a manageable mathematical structure. Every model simplifies, so strong learners need to understand both usefulness and limitation.
Choose variables that capture the problem
A taxi fare can be modelled with a starting fee plus rate per kilometre. A population may be approximated by exponential growth over a limited interval. The variable choice determines what the model can answer.
State assumptions when they matter
A constant speed model assumes speed does not vary. A linear cost model assumes the per-unit rate remains constant. A geometric model may ignore material thickness.
Formulate before calculating
Write the relationship first. Calculator-first behaviour can obscure a bad model.
Interpret residual mismatch
If actual data deviate from the model, the difference may reveal noise, omitted variables or a model that works only within a limited range.
Models can support decisions without being perfect
A rough model can still be useful if its limitations are known. Mathematical judgement includes deciding whether the approximation is good enough for the purpose.
Use sensitivity questions
What happens if one assumption changes? If price increases 10%, if speed falls, if sample size doubles? Sensitivity builds understanding of the structure rather than one answer.
Recovery evidence
A stronger learner can formulate a model, use it, return the result to context and name one important assumption or limitation when relevant.
16. Reason and communicate mathematically
Mathematical reasoning is the ability to connect statements so that another person can inspect why the conclusion follows. A correct final answer can still be weak evidence if the argument is missing.
Justification depends on the claim
If the task asks why two triangles are similar, state the relevant angle or side relationships. If it asks why a root is rejected, connect the rejection to context.
Use mathematical sentences
Symbols are efficient, but explanatory words often reveal the logic between them: “Since the angles are equal…”, “Therefore…”, “Because the quantity must be positive…”.
Do not explain routine arithmetic unnecessarily
Reasoning is not maximum prose. Communicate the mathematical relationship that is not obvious from the working.
Use counterexamples
To disprove a universal statement, one valid counterexample can be enough. This develops logical precision.
Distinguish example from proof
Testing x=2 and x=3 may support a pattern but does not prove it for all x. Learners should know when a general argument is required.
Write assumptions explicitly where needed
Models and geometric arguments can depend on conditions. State them so the reader knows what the reasoning rests on.
Recovery evidence
A stronger learner can produce concise mathematical arguments and knows when a result needs justification rather than just calculation.
17. Move between words, tables, graphs and equations
Representation switching is one of the clearest markers of mathematical maturity. The learner sees the same relationship in several forms and chooses the one that makes the problem easiest.
Words to equations
Translate relationships, not individual words. “Three more than twice x” becomes 2x+3 because the structure is multiplicative then additive.
Tables to graphs
Ordered pairs reveal pattern visually. A constant first difference can suggest linear structure; changing differences may suggest another relationship.
Graphs to equations
Gradient and intercept can recover a linear equation. Intersections can represent simultaneous solutions.
Equations to diagrams
Geometry and coordinate problems can become clearer when algebraic conditions are sketched.
Use representation to reduce memory load
A tree diagram stores conditional probability structure outside working memory. A labelled diagram stores spatial constraints. Good representation is a cognitive tool.
Do not stay loyal to the first representation
If algebra becomes messy, a graph may reveal structure. If a graph is imprecise, algebra may give exact values.
Recovery evidence
A stronger learner chooses representation strategically rather than because the question format suggests it.
18. Check without repeating the same mistake
“Check your work” is poor advice unless the learner has a different checking method. Re-reading the same calculation often reproduces the same assumption.
Estimate
Use scale and sign to reject impossible results quickly.
Substitute
Put a solved value back into the original equation or relationship.
Use an alternative method
Check simultaneous equations graphically, or solve a geometry result through another relationship when time permits.
Check units
If a rate is reported in square centimetres, something is wrong.
Check conditions
Was the denominator allowed to be zero? Is the root within the domain? Is the length positive?
Check interpretation
A mathematical value can be valid but impractical. Round up buses, reject negative people, interpret probability between 0 and 1.
Use targeted checks under time pressure
Not every line can be recomputed. Prioritise multi-step problems, fragile algebra and answers that feel inconsistent with scale.
Recovery evidence
A stronger learner catches errors using independent evidence rather than simply repeating the original process.
19. Timing and method choice
Mathematics timing is often a method-selection problem rather than a writing-speed problem.
Slow recognition delays everything
If Alicia spends two minutes deciding whether a problem is proportional or algebraic, the calculation may be fast but the question still consumes time.
Overcomplicated methods create hidden cost
Tricia solves a simple linear relation through a long algebraic detour. The method is valid but inefficient.
Calculator searching wastes time
Kai Kai repeatedly tries unfamiliar calculator functions rather than first deciding the mathematical structure.
Use method fluency before full papers
Practise choosing methods from mixed short questions. The target is decision speed, not only final correctness.
Protect reasoning marks
Do not rush explanations so aggressively that the mathematical argument disappears.
Use return strategy for stuck problems
If one problem blocks progress, record a meaningful start—diagram, equation, known relationship—then move on if examination conditions permit.
Build stamina gradually
Short mixed sets can stabilise selection before full-paper timing is added.
Recovery evidence
A stronger learner reaches more of the paper because method recognition and execution have become more economical, not merely because handwriting is faster.
20. Transfer across unfamiliar contexts
Transfer is the point at which Mathematics stops being tied to worksheet layout. The learner recognises structure despite new surface details.
Change the story, keep the mathematics
A simultaneous-equations structure can appear in ticket sales, mixtures, ages or purchases. The nouns change; the relationship remains.
Change the representation
Present the same relationship as a table, graph and verbal scenario.
Delay the retry
A problem solved correctly five minutes after modelling may reflect memory of the method. Return a week later in a mixed set.
Mix topics
Blocked practice is useful during initial learning. Mixed practice becomes essential when the learner must select methods independently.
Use near and far transfer
Near transfer changes numbers and context slightly. Far transfer combines topics or hides the familiar structure inside a new problem.
Do not confuse novelty with difficulty
A new context should test transfer, not obscure the mathematics with unnecessarily difficult reading.
Recovery evidence
A stronger learner recognises mathematical structure before being told the chapter and can justify the chosen method in unfamiliar settings.
PG3 Mathematics transfer bank: forty problems that test the mechanism, not the chapter label
This bank is original teaching material for delayed and mixed practice. It is designed to reveal whether a PG3 learner can recognise mathematical structure, choose a representation, execute accurately, interpret the result and check it without being told the chapter. It is not an official SEC paper, specimen-paper reproduction or private placement test. Use the tasks selectively rather than turning the bank into another volume target.
Task 1 · Percentage reversal
A jacket price rises by 25% and later falls by 25%. Is the final price the same as the original? Show the relationship rather than guessing.
Discussion
Let the original price be P. After the rise: 1.25P. After the fall: 0.75×1.25P=0.9375P. The final price is 93.75% of the original, a 6.25% net decrease. Equal percentage changes in opposite directions do not cancel because the second percentage acts on a different base.
Task 2 · Percentage points versus relative percentage
A survey response rises from 40% to 50%. State the increase in percentage points and the relative percentage increase.
Discussion
The increase is 10 percentage points. Relative to the original 40%, the increase is 10/40=25%.
Task 3 · Estimate before exact calculation
Without a calculator, estimate 19.7×49.8. Then explain what an exact calculator answer of 9800 would tell you.
Discussion
Estimate 20×50≈1000. An output near 9800 would be about ten times too large and should trigger an input or decimal-place check.
Task 4 · Fraction structure
Simplify (3/4)x + (5/8)x and explain why converting immediately to decimals is not always the strongest route.
Discussion
3/4=6/8, so the expression is 11/8 x. Exact fractional form preserves exactness and may combine more naturally with later algebra.
Task 5 · Define the variable
A rectangle has length 7 m more than width and area 144 m². Form an equation before solving.
Discussion
Let width=w metres; length=w+7. Then w(w+7)=144, so w²+7w−144=0. The important first act is the model, not the factorisation.
Task 6 · Interpret quadratic roots
Suppose Task 5 produces roots w=9 and w=−16. Which root survives, and why?
Discussion
w=9 survives because width is a physical length and must be positive. The negative algebraic root is rejected in context.
Task 7 · Translation trap
Translate: “The larger number is five less than three times the smaller number.” Let smaller=s and larger=L.
Discussion
L=3s−5. The wording “five less than three times” means subtract 5 after multiplying by 3.
Task 8 · Simultaneous equations from context
A cinema sells 60 tickets. Adult tickets cost $12, student tickets $8, and total revenue is $600. Form and solve the equations.
Discussion
Let a adults and s students. a+s=60; 12a+8s=600. Substitute s=60−a: 12a+480−8a=600, so 4a=120, a=30 and s=30.
Task 9 · Detect impossible contextual data
A ticket problem produces a=27.5 adults. What should the learner do?
Discussion
Do not force a rounded answer blindly. Whole people cannot produce 27.5 adult tickets under the stated model. Recheck the equations and data; if both are correct, the given data are inconsistent with whole counts.
Task 10 · Inequality language
A hall can hold at most 240 people. There are already 78 people inside. Write an inequality for the number x of additional people allowed.
Discussion
78+x≤240, so x≤162.
Task 11 · Direct proportion
y is directly proportional to x. When x=6, y=15. Find y when x=14.
Discussion
y=kx, so k=15/6=2.5. When x=14, y=35.
Task 12 · Inverse proportion
For a fixed journey distance, travel time t is inversely proportional to constant speed v. If t=4 hours at 60 km/h, find the time at 80 km/h.
Discussion
tv is constant: 4×60=240. t=240/80=3 hours.
Task 13 · Unit rate comparison
Pack A costs $8.40 for 6 units. Pack B costs $10.50 for 8 units. Which has the lower unit price?
Discussion
A: $1.40 per unit. B: $1.3125 per unit. Pack B has the lower unit price.
Task 14 · Linear cost intersection
Plan A charges $18 plus $4 per session. Plan B charges $7 per session. When are the costs equal?
Discussion
18+4x=7x, so 18=3x and x=6 sessions. Before 6, B is cheaper; after 6, A becomes cheaper under the model.
Task 15 · Gradient meaning
A distance–time graph is linear from (0,0) to (5,30), where time is hours and distance kilometres. Find and interpret the gradient.
Discussion
Gradient=30/5=6 km/h. It represents constant speed over the interval.
Task 16 · Intercept meaning
A taxi cost graph has equation C=3.2d+5. Interpret 3.2 and 5.
Discussion
3.2 is the cost per unit distance (depending on units in the question); 5 is the starting/base fee when d=0.
Task 17 · Graph versus direct proportion
A straight line has equation y=4x+3. Is y directly proportional to x?
Discussion
No. Direct proportion requires y=kx and a graph through the origin. The +3 intercept breaks direct proportionality.
Task 18 · Quadratic roots from factorisation
Solve x²−7x+12=0 and connect the roots to the graph.
Discussion
(x−3)(x−4)=0, so x=3 or 4. These are the x-intercepts of y=x²−7x+12.
Task 19 · Coordinate midpoint
Find the midpoint of A(−2,5) and B(8,−1).
Discussion
((−2+8)/2,(5−1)/2)=(3,2).
Task 20 · Coordinate gradient and parallelism
Line P passes through (1,2) and (5,10). Line Q has gradient 2. Are they parallel?
Discussion
Gradient of P=(10−2)/(5−1)=8/4=2. Equal gradients mean the distinct lines are parallel.
Task 21 · Geometry scale factor
Two similar shapes have linear scale factor 3 from small to large. If the small area is 12 cm², find the large area.
Discussion
Area factor=3²=9. Large area=108 cm².
Task 22 · Volume scale factor
Using the same linear scale factor 3, a small solid has volume 20 cm³. Find the large volume.
Discussion
Volume factor=3³=27. Large volume=540 cm³.
Task 23 · Pythagoras condition
A triangle has sides 5 cm, 7 cm and 9 cm. Can Pythagoras be used directly to find an unknown side?
Discussion
Not unless a right angle is known or established. Pythagoras is a relationship for right-angled triangles; visual appearance is insufficient.
Task 24 · Trigonometric choice
In a right triangle relative to angle θ, the opposite side is 8 and adjacent side is 15. Which ratio directly gives θ?
Discussion
tan θ=8/15, so θ=tan⁻¹(8/15).
Task 25 · Bearing discipline
A point B lies on a bearing 065° from A. Explain how the angle is measured.
Discussion
Measure 65° clockwise from north at A. The three-figure bearing notation keeps the directional convention explicit.
Task 26 · Mean versus median
Data: 10,11,11,12,12,13,14,40. Find mean and median, then say which may better represent a typical value.
Discussion
Sum=123, mean=15.375. Median=(12+12)/2=12. The extreme value 40 pulls the mean upward, so median may better represent a typical observation.
Task 27 · Same mean, different spread
Set A: 48,49,50,51,52. Set B: 30,40,50,60,70. Both have mean 50. What important difference remains?
Discussion
Set B has far greater spread/variability. Centre alone does not describe distribution.
Task 28 · Sampling claim
A survey of 25 students in one sports CCA finds 80% prefer longer PE lessons. Can the result represent every student in the school?
Discussion
Not reliably. The sample is small and selected from a group likely to have particular interest in physical activity, so representativeness is limited.
Task 29 · Correlation caution
A dataset shows students who spend more time reading tend to have larger vocabularies. Does this prove reading time alone caused the difference?
Discussion
No. The association may reflect other variables such as prior language exposure, motivation or education. Correlation alone does not establish causation.
Task 30 · Basic probability
A bag contains 5 red, 3 blue and 2 green tokens. Find P(blue).
Discussion
There are 10 tokens, so P(blue)=3/10.
Task 31 · Without replacement
Using the same bag, find P(red then blue without replacement).
Discussion
5/10×3/9=15/90=1/6.
Task 32 · With replacement
Find P(red then blue if the first token is replaced before the second draw).
Discussion
5/10×3/10=15/100=3/20.
Task 33 · Complement strategy
A fair coin is tossed three times. Find the probability of at least one head.
Discussion
P(no heads)=P(TTT)=1/8. Therefore P(at least one head)=1−1/8=7/8.
Task 34 · Model assumption
A parking fee is modelled by C=2+1.5h for all h. Name one assumption that could fail in real life.
Discussion
The model assumes the hourly rate stays constant and no daily cap, stepped pricing or extra charge applies. Any relevant stated assumption works.
Task 35 · Sensitivity
Using C=2+1.5h, how much does cost change if parking time rises by 4 hours?
Discussion
The change is 1.5×4=$6. The base fee is unchanged, so it does not affect the difference.
Task 36 · Counterexample reasoning
A student claims “If a number is divisible by 4, it is divisible by 8.” Give a counterexample.
Discussion
12 is divisible by 4 but not by 8. One counterexample disproves the universal statement.
Task 37 · Example versus proof
A learner tests n=1,2,3 and observes n²+n is even. Has the learner proved it for every integer n?
Discussion
No. Examples support the conjecture but do not prove it universally. A general argument can note n²+n=n(n+1), the product of two consecutive integers, one of which is always even.
Task 38 · Checking by substitution
A learner claims x=7 solves 3x−4=18. Check independently.
Discussion
3(7)−4=21−4=17, not 18. The solution is incorrect. Solving gives 3x=22, x=22/3.
Task 39 · Checking contextual reasonableness
A calculation gives 3.2 buses needed for a trip. What answer should be reported if every passenger must have transport?
Discussion
4 buses. Context determines that the discrete practical requirement must be rounded up, not to the nearest whole number.
Task 40 · Independence transfer
One week later, present four unseen problems: one requiring equation formulation, one graph interpretation, one geometry theorem choice and one probability sample space. Remove chapter headings and do not tell the learner which method is relevant. Record the first thirty seconds: what concept, representation and method does the learner select independently?
This final task is the core of the bank. The purpose of the earlier thirty-nine is to build mathematical relationships; the purpose of the delayed mixed task is to discover whether the learner owns the selection process.
How to use this bank
If algebraic execution is the active bottleneck, use Tasks 5–10 and 18 before returning to mixed work. If representation is weak, use Tasks 5, 8, 14–20, 23–25 and 34. If statistical judgement is weak, use Tasks 26–29. If checking is weak, use Tasks 3, 6, 9, 17, 23, 38 and 39. If method selection is the bottleneck, do not repeatedly solve the whole bank in order; mix questions from different families and ask for method choice before execution.
The bank should become less necessary as mathematical independence grows. A good practice bank is eventually replaced by the learner’s ability to recognise structure in problems they have never seen before.
PG3 Mathematics mechanism clinics: when the obvious diagnosis is wrong
Mathematics errors are frequently labelled with broad words—careless, weak algebra, slow, does not read, forgets formulas. Those labels can describe the surface without identifying the repair. The clinics below separate errors that look similar in the final answer but begin at different points in the mathematical engine.
Clinic 1 · “Careless sign error” that is really bracket structure
A learner expands −3(x−4) as −3x−12. The error is often called carelessness, but if it recurs whenever a negative factor meets a bracket, the problem is structural. The learner is not distributing multiplication across every term while preserving sign.
Repair: slow the structure down: −3(x−4)=−3×x+(−3)×(−4)=−3x+12. Then vary the coefficient, bracket order and context. The target is not “be more careful”; it is “see the bracket as one multiplication applied to two terms”.
Clinic 2 · “Weak algebra” that is really variable definition
A learner forms equations inconsistently because x changes meaning halfway through the solution. In one line x represents the adult ticket price; later it silently becomes the number of adult tickets.
Repair: define variables with quantity and unit before forming equations. If the learner cannot say what x means in words at any stage, the representation has become unstable.
Clinic 3 · Correct equation, wrong interpretation
A quadratic model produces roots 8 and −12. The learner reports both because the algebra is correct.
Repair: separate mathematical solution from contextual solution. After solving, ask: what quantity does the variable represent, what values are physically or logically permitted, and which roots satisfy those conditions?
Clinic 4 · “Doesn’t understand percentages” that is really changing base
A student expects a 20% rise followed by a 20% fall to return to the original amount.
Repair: express each change as a multiplier. The second percentage acts on the already-changed value. Use 100→120→96 before returning to algebraic P×1.2×0.8.
Clinic 5 · “Doesn’t read the question” that is really scope language
A learner solves for “at least 12” as x=12 rather than x≥12.
Repair: build a mathematical language dictionary around at least, at most, no more than, greater than, between, inclusive, consecutive, per, difference and total. Then use those words in mixed contexts so they become structural cues rather than vocabulary trivia.
Clinic 6 · Slow simultaneous equations because method is not selected
Tricia always uses substitution, even when one equation is 3x+2y=17 and another is 3x−5y=3, where elimination is immediate.
Repair: before solving, compare coefficients and state which variable can be eliminated most cheaply. Practise choosing the method without completing the whole calculation.
Clinic 7 · Fast simultaneous equations with weak checking
Kai Kai obtains x=4, y=2 quickly but never substitutes into both original equations.
Repair: build one independent check into the routine. For a pair of equations, substitute the pair into both. The check should not merely repeat the elimination arithmetic.
Clinic 8 · Graph plotting is accurate but the relationship is not understood
A learner plots y=2x+5 perfectly but cannot explain what 2 and 5 mean.
Repair: interpret gradient as change in y per unit change in x, and the intercept as the value when x=0. Then use contextual equations—cost, distance, temperature—to make the parameters meaningful.
Clinic 9 · Straight line mistaken for direct proportion
The learner sees a straight graph and automatically writes y∝x.
Repair: direct proportion requires y=kx and therefore passage through the origin. Contrast y=3x with y=3x+7 using tables, equations and graphs.
Clinic 10 · Gradient calculation without dimensional meaning
The learner calculates (15−5)/(8−3)=2 but cannot state units.
Repair: write “vertical quantity per horizontal quantity” before arithmetic. Units often reveal whether the chosen points or interpretation make sense.
Clinic 11 · Quadratic factorisation memorised but representation rigid
A learner tries to factorise every quadratic, spending too long when integer factors are not convenient.
Repair: compare factorisation, formula and graphical methods. Method choice depends on the coefficients, required exactness and context, not on loyalty to one technique.
Clinic 12 · Completing the square learned as a ritual
The learner can perform the algebra but does not see why the form a(x−h)²+k is useful.
Repair: connect the completed-square form to the turning point and graph translation. The transformation is a representation choice, not merely another expansion exercise.
Clinic 13 · Geometry theorem chosen from picture shape
A triangle looks right-angled, so the learner uses Pythagoras even though no right angle is stated or proved.
Repair: require a condition statement before theorem use: “This triangle is right-angled because… therefore Pythagoras applies.” If the condition cannot be stated, theorem choice is not justified.
Clinic 14 · Similarity confused with congruence
A learner treats corresponding sides in similar figures as equal rather than proportional.
Repair: contrast the invariants. Congruent figures have equal corresponding lengths; similar figures preserve angles and a constant scale factor. Use linear, area and volume scale factors together.
Clinic 15 · Trigonometry button-searching
The learner tries sin, cos and tan until one gives a plausible number.
Repair: label the known sides relative to the chosen angle, then select the ratio that contains exactly the needed quantities. The calculator comes after the geometry.
Clinic 16 · Degrees/radians or calculator mode ignored
A sensible setup produces an impossible angle because the calculator is in the wrong mode.
Repair: include mode and reasonableness in the checking routine. An acute angle in a right triangle should not emerge as hundreds of degrees.
Clinic 17 · 3D geometry overwhelms because the hidden 2D triangle is unseen
A learner stares at a cuboid and cannot decide how to begin a space-diagonal problem.
Repair: identify the face diagonal first, then the second right triangle containing the space diagonal. Complex spatial problems can often be decomposed into familiar 2D structures.
Clinic 18 · Mean used automatically when an outlier dominates
Data include one extreme value, yet the learner reports mean as “the average” without comment.
Repair: compare mean and median and discuss how the outlier changes each. The question is not which measure is universally better, but which summary is informative for this dataset and purpose.
Clinic 19 · Graph scale creates visual exaggeration
Two bars differ from 98 to 100, but the axis begins at 97 and the learner describes the difference as enormous.
Repair: read numerical magnitude separately from visual magnitude. A truncated scale can be legitimate for detail but should not distort interpretation.
Clinic 20 · Correlation treated as causation
Students with more tuition hours have higher marks in a sample, so the learner concludes tuition hours caused the improvement.
Repair: generate alternative explanations: prior attainment, motivation, family resources, subject difficulty or self-selection. Observational association does not isolate causation.
Clinic 21 · Probability denominator stays frozen after an event
Without replacement, the learner uses the original total for the second draw.
Repair: redraw the sample space after the first event. The denominator changes because one object has been removed.
Clinic 22 · “At least one” counted by incomplete listing
The learner manually lists several outcomes and misses one.
Repair: test whether the complement “none” is simpler. Strategic representation reduces counting error.
Clinic 23 · Model gives a decimal answer to a discrete decision
A bus calculation gives 4.1 buses and the learner rounds to 4.
Repair: interpretation follows the practical constraint. If every passenger must travel, 5 buses are required. Mathematical rounding conventions do not override the real-world requirement.
Clinic 24 · Model assumed valid beyond its range
A linear cost equation fitted to the first ten hours is extended to 100 hours even though the venue has a daily cap.
Repair: state domain and assumptions. A model can be mathematically correct and practically invalid outside the conditions for which it was built.
Clinic 25 · Correct examples mistaken for proof
A learner checks a statement for n=1,2,3,4 and declares it true for all integers.
Repair: distinguish inductive evidence from deductive argument. Search for structure—factorisation, parity, inequalities or counterexample—depending on the claim.
Clinic 26 · One counterexample overlooked
A universal claim fails for one valid case, but the learner says “most examples work”.
Repair: universal statements are disproved by a single counterexample. This is a logical rule, not a matter of majority.
Clinic 27 · Exact and approximate answers mixed carelessly
A learner rounds an intermediate trigonometric value and uses it through several later steps.
Repair: retain full calculator precision or exact form through working, then round once at the end unless the task requires otherwise.
Clinic 28 · Unit conversion hidden inside algebra
A rate uses kilometres and seconds, producing a number that looks plausible but has mixed units.
Repair: standardise units before forming or interpreting the relationship. Units are part of the mathematics, not labels added after calculation.
Clinic 29 · The learner “checks” by doing the same calculation again
The same wrong input is entered twice, giving the same wrong answer and false confidence.
Repair: use a different checking channel: estimation, inverse operation, substitution, graph, unit analysis or alternative method.
Clinic 30 · Timed work fails only after one difficult question
A learner spends seven extra minutes on one unfamiliar problem, then rushes three routine questions.
Repair: build a containment routine: preserve useful working, mark the question, re-enter the paper, return only if expected value later justifies it. Timing is a resource-allocation problem.
What these clinics change
The purpose is not to replace one giant syllabus with thirty new labels. It is to make the first weak mathematical decision visible. If “careless” is actually unstable bracket structure, teach brackets. If “slow” is actually method indecision, train selection. If “weak problem solving” is actually representation rigidity, practise switching representations.
Once the true bottleneck is named, practice becomes smaller, more precise and easier to verify on changed problems.
Integrated G3 Mathematics workbook: twenty-five problems where topics have to work together
The earlier transfer bank isolates individual relationships. This workbook deliberately removes that protection. Each problem combines two or more mathematical jobs so the learner has to decide what matters before calculation begins. The problems are original teaching material, not official SEC questions, specimen-paper reproductions or grade predictors.
Use them after the underlying procedures are reasonably secure. For each problem, ask the learner to write three short lines before solving: what is given; what is required; what representation or method is likely to help. After solving, add one independent check or interpretation. The pre-solution and post-solution decisions are part of the Mathematics.
Problem 1 · Transport pricing, equations and discrete interpretation
A school hires minibuses for a trip. Company A charges a fixed booking fee of $96 plus $42 per bus. Company B charges no booking fee but $58 per bus. Each bus carries at most 18 students. There are 143 students.
(a) Find the number of buses required. (b) Write cost models for both companies. (c) Which company is cheaper for the required number of buses? (d) Find the theoretical number of buses at which the two models have equal cost and explain why that value may not itself be a practical booking choice.
Worked discussion
(a) 143/18≈7.94, so 8 buses are required because every student must have a place. (b) Let b be the number of buses: A=96+42b; B=58b. (c) At b=8, A=96+336=$432; B=464, so A is cheaper. (d) 96+42b=58b gives 96=16b, b=6. The equality happens at exactly 6 here, which is a valid whole number; if a different model produced a fraction, the learner would compare nearby whole booking quantities rather than booking a fraction of a bus. The deeper point is that discrete context controls interpretation.
Problem 2 · Rectangle, algebra and percentage area change
A rectangular garden has width w metres and length w+5 metres. Its area is 84 m².
(a) Form and solve an equation for w. (b) A one-metre-wide path is added outside all four sides. Find the new total area. (c) Calculate the percentage increase in total area.
Worked discussion
(a) w(w+5)=84, so w²+5w−84=0=(w+12)(w−7). Width=7 m and length=12 m; reject −12 as a physical width. (b) Adding 1 m outside each side increases each dimension by 2: new dimensions 9 by 14, area=126 m². (c) Increase=42 m²; percentage increase=42/84×100%=50%.
Problem 3 · Linear graph, intercept and break-even reasoning
A printing shop models cost C dollars for x posters by C=18+1.6x. Another shop uses C=2.2x.
(a) Interpret the gradient and intercept of the first model. (b) Find the break-even quantity. (c) Decide which shop is cheaper for 20 posters and for 50 posters. (d) Sketch the two lines and explain how the graph encodes the same decision.
Worked discussion
(a) $1.60 is the additional cost per poster; $18 is the base/setup cost. (b) 18+1.6x=2.2x gives 18=0.6x, so x=30. (c) x=20: first=50, second=44, so second cheaper. x=50: first=98, second=110, so first cheaper. (d) The lines intersect at x=30. Before that point the no-base-fee line lies lower; after it the lower gradient of the first shop dominates.
Problem 4 · Similarity, area and cost model
A logo is enlarged with linear scale factor 2.5. The original printed area is 48 cm². Printing costs $0.04 per cm² plus a fixed setup fee of $3.
(a) Find the enlarged area. (b) Find the printing cost. (c) Explain why multiplying the original area by 2.5 would be structurally wrong.
Worked discussion
(a) Area scale factor=2.5²=6.25; enlarged area=300 cm². (b) Cost=3+0.04(300)=$15. (c) Area depends on two linear dimensions, so it scales with the square of the linear scale factor.
Problem 5 · Speed, graph interpretation and average rate
A cyclist travels 12 km in the first 30 minutes, rests for 15 minutes, then travels another 18 km in 45 minutes.
(a) Find the cycling speed in each moving segment. (b) Find the average speed for the entire journey including rest. (c) Describe the shape of a distance–time graph. (d) Explain the difference between segment gradient and overall average speed.
Worked discussion
First segment: 12/0.5=24 km/h. Second moving segment: 18/0.75=24 km/h. Total distance=30 km. Total time=0.5+0.25+0.75=1.5 h. Overall average speed=20 km/h. The graph rises linearly, is horizontal during rest, then rises again with the same moving gradient. Segment gradient measures local constant speed; overall average includes the stationary interval.
Problem 6 · Statistics, outlier and percentage comparison
Two revision groups record minutes taken to complete the same mixed quiz.
Group A: 28, 29, 30, 30, 31, 32, 33, 47.
Group B: 27, 28, 29, 30, 31, 32, 33, 34.
(a) Find the mean and median for each. (b) Compare typical performance and spread. (c) Explain why the single value 47 matters to interpretation. (d) If Group A’s 47-minute result came from an interrupted session, what would you do before simply deleting it?
Worked discussion
A sum=260, mean=32.5, median=(30+31)/2=30.5. B sum=244, mean=30.5, median=30.5. Both have the same median, but A’s mean and range are raised by 47. Before deleting 47, verify the interruption and decide whether the value belongs to the intended dataset. If the session was genuinely not comparable, exclusion may be defensible if documented; automatic deletion merely because the value is inconvenient is not.
Problem 7 · Probability, conditional sample space and expected interpretation
A box contains 6 red, 4 blue and 2 green counters. Two counters are drawn without replacement.
(a) Find P(two red). (b) Find P(at least one green) using a complement. (c) Explain how the second-draw denominator changes after the first draw.
Worked discussion
(a) 6/12×5/11=30/132=5/22. (b) P(no green)=10/12×9/11=90/132=15/22, so P(at least one green)=7/22. (c) Without replacement, one counter has been removed, so the sample space falls from 12 to 11 and the colour counts may also change.
Problem 8 · Coordinate geometry and modelling
A delivery route is approximated by a straight line between A(2,3) and B(10,15) on a map grid where each coordinate unit represents 0.5 km.
(a) Find the gradient. (b) Find the midpoint in grid coordinates. (c) Find the straight-line distance in kilometres. (d) State one limitation of using this distance as actual travel distance.
Worked discussion
(a) Gradient=(15−3)/(10−2)=12/8=1.5. (b) Midpoint=(6,9). (c) Grid distance=√(8²+12²)=√208=4√13≈14.42 units; physical distance≈7.21 km. (d) Actual roads may not follow a straight line and can include turns, restrictions or terrain.
Problem 9 · Inequality, budget and integer constraint
A club has a budget of $500. Venue hire costs $140 and each participant pack costs $18.
(a) Form an inequality for the maximum number n of packs. (b) Find the maximum whole number of participants. (c) Explain why simply solving the equality is not the entire job.
Worked discussion
140+18n≤500. Then 18n≤360, n≤20. Maximum is 20 packs. Equality happens exactly here, but generally an inequality asks for a permissible range and the context may require a whole-number maximum below a non-integer bound.
Problem 10 · Proportion and inverse reasoning
Six identical machines complete a job in 15 hours under an idealised model where total machine-hours are constant.
(a) How long would 10 identical machines take? (b) State the assumption behind the calculation. (c) Give one real-world reason the inverse-proportion model might fail.
Worked discussion
Total machine-hours=6×15=90. Ten machines require 90/10=9 hours. The model assumes equal machines, constant productivity and no interference or setup loss. In reality, crowding, resource limits or coordination can reduce efficiency.
Problem 11 · Trigonometry, exact planning and interpretation
A ladder 6.5 m long reaches a wall. Its foot is 2.4 m from the wall.
(a) Find the height reached. (b) Find the angle the ladder makes with the ground. (c) Give one independent check on the height.
Worked discussion
Height=√(6.5²−2.4²)=√(42.25−5.76)=√36.49≈6.04 m. cos θ=2.4/6.5, so θ≈68.3°. A check: the height must be less than 6.5 m and greater than √(6.5²−3²), so 6.04 is plausible; alternatively substitute the legs back into Pythagoras.
Problem 12 · Bearings and triangle reasoning
From point A, B is on a bearing 060° and C is due east of A. Explain the angle BAC before any distance calculation.
Worked discussion
Due east is bearing 090°. The angle between AB at 060° and AC at 090° is 30°. Drawing north and measuring clockwise prevents orientation mistakes.
Problem 13 · Quadratic model and maximum interpretation
A simplified profit model is P=−2x²+40x−120, where x is the number of tens of units sold.
(a) Find the x-coordinate of the maximum using completing the square or the vertex formula. (b) Find the maximum modelled profit. (c) Explain one reason not to assume the quadratic model remains valid for arbitrarily large x.
Worked discussion
x=−b/(2a)=−40/(−4)=10. P(10)=−200+400−120=80. The model may only approximate a certain sales range; costs, capacity or pricing can change outside it.
Problem 14 · Algebra and geometry combined
A right triangle has shorter legs x and x+7 and hypotenuse 17.
(a) Form an equation. (b) Solve for x. (c) Check the dimensions.
Worked discussion
x²+(x+7)²=17². Then 2x²+14x+49=289, so x²+7x−120=0=(x+15)(x−8). x=8 (reject −15). The legs are 8 and 15; 8²+15²=64+225=289=17².
Problem 15 · Data representation and misleading visual impression
School A’s pass rate rises from 92% to 94%. School B’s rises from 60% to 66%.
(a) Find the percentage-point increase for each. (b) Find each relative percentage increase. (c) Explain why a chart using a vertical axis from 90% to 95% for School A alone could visually exaggerate the size of its change.
Worked discussion
A: +2 percentage points; relative increase 2/92≈2.17%. B: +6 percentage points; relative increase 6/60=10%. A truncated axis can make a 2-point change occupy most of the graph height, so visual difference must be read with the numerical scale.
Problem 16 · Probability and expected counts
A game has probability 0.3 of a win on each independent play. Over 200 plays, what is the expected number of wins? Does this mean exactly that many wins must occur?
Worked discussion
Expected wins=0.3×200=60. This is a long-run average expectation, not a guarantee that exactly 60 wins will occur in one set of 200 plays.
Problem 17 · Formula rearrangement and units
Density ρ=m/V. A material has density 2.7 g/cm³ and volume 35 cm³.
(a) Find mass. (b) If volume is instead given as 0.035 dm³, explain what must happen before substitution.
Worked discussion
m=ρV=2.7×35=94.5 g. The units must be made compatible. 1 dm³=1000 cm³, so 0.035 dm³=35 cm³ before using 2.7 g/cm³.
Problem 18 · Sequence reasoning
A sequence begins 5, 9, 13, 17, …
(a) Find the nth term. (b) Is 202 a term? (c) Explain the reasoning.
Worked discussion
nth term=4n+1. Set 4n+1=202 gives n=201/4=50.25, not a positive integer, so 202 is not a term. The integer requirement matters.
Problem 19 · Algebraic identity versus numerical evidence
Show that the difference between squares of consecutive integers is always odd.
Worked discussion
Let consecutive integers be n and n+1. (n+1)²−n²=n²+2n+1−n²=2n+1, which is odd for every integer n. Testing examples may suggest the pattern, but the algebra proves it generally.
Problem 20 · Counterexample and domain
Claim: “If ab=0, then a=0 and b=0.” Is the claim true?
Worked discussion
No. If a=0 and b=5, then ab=0 while b≠0. The correct zero-product statement is that at least one factor is zero.
Problem 21 · Multi-stage percentage and financial interpretation
An item costs $240 before a 15% discount, followed by 9% tax on the discounted price.
(a) Find the final price. (b) Explain why subtracting 15% and adding 9% to get a “net 6% discount” is not generally exact.
Worked discussion
Discounted price=240×0.85=$204. Taxed price=204×1.09=$222.36. The two percentages act on different bases, so the combined multiplier is 0.85×1.09=0.9265, a 7.35% net reduction from the original.
Problem 22 · Histogram/data judgement without overclaim
Two classes have the same median mark but Class X has a much wider interquartile range. What can be concluded?
Worked discussion
The typical central position by median is similar, but the middle 50% of Class X’s marks are more spread out. This suggests greater variability in Class X. It does not by itself explain why.
Problem 23 · Model selection from words
A tank begins with 500 litres and loses 12 litres each minute at a constant rate.
(a) Write a linear model V(t). (b) Find when the model reaches zero. (c) State one domain restriction.
Worked discussion
V=500−12t. Set V=0: t=500/12≈41.67 min. The model should not be extended meaningfully beyond empty volume; practical rate may also change near the end depending on the physical system.
Problem 24 · Mixed checking clinic
A learner obtains x=−3 as the radius of a circle after solving an equation and writes area=9π cm² because r² is positive. Explain the mathematical and contextual problem.
Worked discussion
The algebraic expression r² can produce a positive area from a negative numerical r, but radius as a physical length cannot be negative. The root must be interpreted before substitution into later formulas.
Problem 25 · Timed decision drill
Prepare a mixed set of six unseen questions: one percentage, one simultaneous-equation context, one graph relationship, one geometry problem, one statistical comparison and one probability question. Give the learner only ninety seconds to write the likely representation/method for all six without solving them.
Then remove the time pressure and solve. Compare two records: method-selection accuracy and execution accuracy. If execution is strong but the ninety-second selection record is weak, the timing bottleneck sits before calculation. If selection is strong but execution fails, procedural repair is more relevant.
How to read the integrated workbook
The goal is not twenty-five perfect answers in one sitting. The goal is to observe where mathematical control breaks when multiple mechanisms coexist. A learner who solves routine algebra but fails Problems 1, 2, 8, 14 and 23 may have a formulation/interpretation ceiling. A learner who solves but cannot check Problems 11, 14, 17 and 24 may need independent validation strategies. A learner who understands all problems untimed but fails Problem 25 may need faster structure recognition rather than more content teaching.
Use the evidence to reduce practice, not inflate it. Once the active mechanism is known, choose the smallest set of problems that tests it across changed contexts and delayed attempts.
21. Twelve-week G3 Mathematics build
A twelve-week cycle is long enough to diagnose, teach, mix, transfer and retest. It is an instructional framework, not an official school timeline or grade guarantee.
Weeks 1–2 · Diagnose the engine
Use short probes across number, algebra, graphs, geometry, statistics and problem solving. Classify the first failure: concept, representation, method selection, execution, interpretation, checking or reasoning.
Weeks 3–4 · Repair the highest-leverage mechanism
If representation is weak, practise turning words into equations, diagrams and tables. If algebra execution is weak, stabilise exact procedures. If interpretation is weak, require every answer to return to context.
Weeks 5–6 · Mix nearby topics
Combine percentage with algebra, graphs with rates, geometry with equations. The learner begins selecting rather than following chapter labels.
Weeks 7–8 · Increase unfamiliarity
Use new contexts and multi-step questions while keeping the underlying mathematics within reach.
Weeks 9–10 · Add timed decision-making
Use short mixed sets and record where time is spent. Focus on method selection and checking efficiency.
Week 11 · Integrated problem-solving workshop
Use one scenario requiring algebra, graph interpretation, geometry and data reasoning.
Week 12 · Unseen review
Compare independence, method choice, error families and checking with Week 1. Update the mathematical profile.
Keep one strength alive
Every cycle should maintain secure topics so fluency does not decay while weaknesses are repaired.
22. Three learners, three different bottlenecks
Alicia · Strong procedures, weak formulation
Alicia solves equations accurately when they are given. She struggles to build them from word problems. Her repair focuses on defining variables, translating relationships and sketching structure before solving.
Tricia · Strong understanding, slow method selection
Tricia understands several methods and often chooses a valid but inefficient route. Her repair compares methods explicitly and uses mixed short sets to build recognition speed.
Kai Kai · Fast execution, weak interpretation and checking
Kai Kai reaches answers quickly but accepts impossible roots, early rounding and unit errors. His repair uses estimate–solve–interpret–check as a fixed closing routine.
Same score, different lesson
All three could obtain 68%. One total cannot distinguish formulation, selection and checking.
23. Integrated mathematical workshop: The Community Hall Project
This workshop is original teaching material, not an official SEC question.
Scenario
A community hall charges a base booking fee of $120 plus $18 per hour. A second venue charges no base fee but $30 per hour. A rectangular activity area inside the hall has perimeter 52 m and length 4 m more than width. Attendance over eight sessions is 18, 22, 24, 24, 26, 29, 31 and 42.
Task 1 · Cost model
Let h be hours. Venue A: C=120+18h. Venue B: C=30h. Solve 120+18h=30h, giving 120=12h and h=10. The costs are equal at 10 hours.
Task 2 · Compare costs
At 6 hours, A costs 120+108=$228; B costs $180, so B is cheaper. At 14 hours, A costs 120+252=$372; B costs $420, so A is cheaper.
Task 3 · Geometry model
Let width=w and length=w+4. Perimeter 52 gives 2w+2(w+4)=52, so 4w+8=52, w=11 and length=15. Area=165 m².
Task 4 · Attendance statistics
Sum=216, mean=27, median=(24+26)/2=25, range=42−18=24. The value 42 raises the mean above the median, so the median may better describe a typical session if the large value is unusual.
Task 5 · Probability extension
If a bag contains 5 red, 3 blue and 2 green tokens, P(red)=5/10=1/2. Without replacement, P(red then blue)=5/10 ×3/9=1/6.
Task 6 · Modelling judgement
The venue equations assume hourly rates remain constant and ignore extra charges. The attendance mean assumes each session is comparable. Strong mathematics keeps those assumptions visible.
What the workshop tests
Representation, equation formulation, arithmetic, interpretation, statistics, probability, modelling and checking appear inside one scenario. The learner must choose methods rather than receive a chapter label.
24. Student, parent and tutor routes
Student route
After each error, write the first failure: concept, representation, method, execution, interpretation or checking. Keep only two active targets at a time.
Parent route
Ask which mathematical decision caused the lost marks rather than whether the learner was “careless”. Repeated carelessness should be converted into a mechanism: sign control, unit tracking, rushed interpretation or weak checking.
Tutor route
Teach the first weak link, then transfer it to changed contexts. Record support honestly. A correct solution after the tutor suggests the equation is teaching evidence, not yet independent formulation.
School conversation
Bring representative work and ask whether the private profile matches classroom performance. Formal subject-level arrangements remain with the school.
Frequently asked questions
Is PG3 the same as G3 Mathematics?
No. PG3 is an entry route; G3 is a subject level.
What is the 2027 G3 Mathematics code?
SEAB lists Mathematics as K310 for 2027 G3 school candidates.
Is G3 Mathematics mainly algebra?
No. Algebra is important, but the syllabus also includes Geometry and Measurement, Statistics and Probability, and mathematical processes such as reasoning, communication, application and modelling.
Why can a student be good at routine questions and weak at problem solving?
Routine questions name or strongly cue the method. Problem solving requires interpretation, representation and selection before execution.
Should every learner do full papers constantly?
No. Full papers are useful for integration and timing, but targeted practice is often more efficient when a specific mechanism is weak.
What is the best sign of readiness?
Increasing independence across unfamiliar problems: correct concept selection, useful representation, reliable execution, interpretation and checking with less prompting.
25. Keep Mathematics larger than the answer
G3 Mathematics is not a race to obtain the final number. It is the controlled use of concepts, representations and reasoning to solve problems whose structure may be hidden.
K310 makes this explicit: standard techniques matter, but so do problem solving, application, reasoning and communication. The learner must know Mathematics and know when and how to use it.
Posting Group 3 should remain in proportion. It is an entry route, not a mathematical identity. The educational question is what the learner can now do independently.
For continued navigation, use How Mathematics Works, What Is G1, G2 and G3 Mathematics in Secondary School? and the How X Works Hub.
The useful final question is not “Can I remember the formula?” It is “Can I recognise the relationship, represent it, choose a method, justify it and decide whether the answer makes sense?”
Official sources and scope
Current route facts were checked on 18 September 2026. Official syllabuses and school arrangements can change.
Singapore Examinations and Assessment Board. 2027 G3 syllabuses for school candidates. Mathematics is listed as K310.
SEAB K310 G3 Mathematics syllabus for 2027. Used for the three content strands and assessment-objective framing. The official syllabus gives approximate AO weightings of 45% standard techniques, 40% problem solving in varied contexts and 15% reasoning/communication.
Ministry of Education, Singapore. Full Subject-Based Banding / secondary curriculum. Used for the distinction between Posting Groups and subject levels.
All worked problems, learner profiles and instructional cycles in this article are original teaching material, not official examination questions, specimen papers, placement tests or grade guarantees.