G2 Mathematics K210 error analysis becomes faster when the learner learns to read the shape of a wrong answer. A negative value where only a positive quantity makes sense, an answer exactly ten times too large, a probability above 1, a gradient with the wrong sign, or a final value that violates an integer constraint each points toward a different broken step. The wrong answer itself can be diagnostic evidence.
This fifty-fifth Learner’s Guide develops a Mathematics error-signature system. Instead of restarting every incorrect solution from line one, the learner asks what kind of wrongness appeared, which transition could create that pattern, and what is the smallest reliable check. The method builds on Vol 0045 error triage and Vol 0051 on exact and rounded answers, but it is narrower: diagnose Mathematics from the signature left by the error.
The official 2027 K210 syllabus gives two two-hour papers, requires all Paper 1 questions, and states that omission of essential working will result in loss of marks. See the official K210 syllabus. Readable working also makes error forensics possible: if the route is visible, the learner can locate the first invalid step instead of guessing from the final number.
The central rule: diagnose before you redo
Re-solving can fix a question, but it often hides the cause. If the second attempt is correct, the learner may conclude that the first error was “careless” and learn nothing portable. Diagnosis asks what distinctive feature of the wrong answer points back to the failure.
A useful repair produces two outputs: a corrected answer and a reusable warning sign.
Five layers of a Mathematics error signature
- Magnitude: too large, too small, impossible scale.
- Direction: positive instead of negative, increase instead of decrease, slope sign reversed.
- Type: length instead of area, probability outside its range, decimal when integer feasibility is required.
- Structure: wrong formula family, wrong base, inverted rate, missing term or constraint.
- Presentation: correct mathematics but wrong unit, accuracy, label or final interpretation.
The signature does not prove the cause, but it narrows the search.
Signature: answer ten, hundred or thousand times too large
This often points to a unit-conversion or place-value error rather than a deep concept failure.
Check whether metres and centimetres, kilograms and grams, percentages and decimals, or powers of ten were converted in the correct direction. Then inspect only the conversion step before rebuilding the rest.
Signature: answer ten, hundred or thousand times too small
The same families of error can operate in reverse: division used where multiplication was needed, decimal point shifted the wrong way, or a percentage treated as a whole number.
Compare the expected scale before touching the algebra. Order-of-magnitude checks can localise the problem in seconds.
Signature: negative value for a physical length or count
A negative result may be algebraically valid but contextually impossible.
Check sign movement, equation setup and whether the negative solution should be rejected by the domain. Do not automatically change the sign to positive; find why the negative appeared.
Signature: positive answer where the trend should decrease
A positive numerical result may contradict the direction described in the question.
Check whether change was defined as final minus initial or initial minus final, and whether the requested quantity was decrease, difference or signed change.
Signature: answer exceeds a clear maximum
If a probability is above 1, a percentage above a stated total, or a part exceeds the whole, the result violates a boundary.
Inspect denominator choice, complement use, double counting and whether a part-to-whole relationship was mistaken for part-to-part.
Signature: probability is negative
Probabilities cannot be negative in the ordinary K210 context.
Check subtraction order, complement reasoning, mutually exclusive event assumptions and whether an algebraic expression was evaluated in a valid range.
Signature: probabilities sum to more than 1
This often signals overlapping cases counted as if they were disjoint, or several outcomes counted twice.
Draw a list, table or simple event structure and identify the overlap rather than merely rescaling the numbers.
Signature: mean lies outside the data range
An ordinary arithmetic mean of a data set should lie between the smallest and largest values.
Check total, number of values, copied data and whether frequency was included correctly. The impossible location of the mean is a strong diagnostic clue.
Signature: median depends on arithmetic when it should not
Learners sometimes average two central values even when the number of observations is odd, or fail to average when it is even.
Count ordered data positions before calculating. A position error is different from an arithmetic error.
Signature: gradient sign contradicts the graph
If the line rises left to right but the calculated gradient is negative, point order may have been mixed inconsistently.
Use the same order for numerator and denominator: change in y over the corresponding change in x. Visual direction is a fast sign check.
Signature: gradient magnitude is implausibly steep or flat
This can come from reading the graph scale incorrectly or reversing axes.
Check one grid interval on each axis before recomputing. A scale error often leaves a clean multiplicative signature.
Signature: intercept is mistaken for gradient
A learner may read a visible coordinate correctly but answer the wrong graph property.
Name the target before calculating: slope, x-intercept, y-intercept, maximum, minimum or coordinate. This is an interpretation failure, not necessarily weak graph knowledge.
Signature: area reported in linear units
The calculation may even contain the correct numbers while the quantity type has been lost.
Trace whether two lengths were multiplied and restore square units. If the formula itself used only one dimension, the error is structural rather than presentational.
Signature: volume reported in square units
A volume requires three-dimensional units.
Check whether the third dimension was included and whether the unit conversion was performed before or after multiplication consistently.
Signature: perimeter and area are numerically confused
A familiar rectangle formula can be applied to the wrong target when the learner sees length and width and calculates automatically.
Ask whether the question measures around or covers a surface. Quantity identity should precede formula retrieval.
Signature: time answer uses the wrong clock system
An elapsed-time calculation can be numerically plausible but attached to the wrong 12-hour or 24-hour interpretation.
Build a short timeline and label morning, afternoon or 24-hour times explicitly. This is often quicker than repeating subtraction.
Signature: rate unit is inverted
If the question asks for kilometres per hour and the answer effectively has hours per kilometre, the ratio was reversed.
Write the unit as a fraction before substituting numbers. The unit reveals numerator and denominator roles.
Signature: unit price grows when quantity increases unexpectedly
The learner may have calculated total cost rather than cost per item, or inverted the division.
Name the rate in words: dollars per item means dollars divided by items. The verbal unit is a formula check.
Signature: percentage change uses the wrong base
A plausible percentage can still be wrong because the denominator is the final value instead of the original, or vice versa.
Write “change ÷ original” or the appropriate defined base before calculating. The issue is modelling, not calculator skill.
Signature: reverse percentage gives a result smaller than the discounted value
If the original price should exceed the sale price but the computed original is smaller, the base relationship is wrong.
Represent sale price as the remaining percentage of original before dividing.
Signature: repeated percentage change is treated as simple addition
Two successive percentage changes act on changing bases.
If the learner simply adds the percentages, the signature may be a result that ignores compounding. Rebuild the multiplicative factors rather than patching the final number.
Signature: ratio answer preserves numbers but not quantity roles
A ratio can be simplified correctly while representing the wrong comparison.
Label each term before simplification. Part A : Part B is not interchangeable with part : whole.
Signature: scale drawing grows in the wrong direction
If a real object should be larger than the drawing but the calculated length is smaller, the scale factor may have been inverted.
Write the correspondence as drawing : actual or actual : drawing and preserve that orientation through the equation.
Signature: map distance is plausible but unit is absurd
A calculation may produce 4.2 and the learner writes kilometres even though the map length was centimetres and no conversion was made.
Keep units attached to scale relationships. The missing conversion is visible when the units do not cancel or transform.
Signature: algebraic solution works in the transformed equation but not the original
Operations such as multiplying by expressions, squaring, or simplifying can introduce or hide restrictions.
Substitute the candidate into the original equation, not only the final rearranged form. Domain checks belong after solving.
Signature: one solution missing from a quadratic-style structure
A learner may take only one square root sign or factor and stop.
When a step produces a squared relationship, ask whether both positive and negative possibilities need consideration within the domain.
Signature: extraneous solution survives
A candidate can satisfy an intermediate transformed equation but fail the original condition.
Substitution is the decisive test. Reject based on the original equation or context, not on appearance.
Signature: equation has the right numbers but wrong operators
This often comes from translating words mechanically without representing the relationship.
Write a verbal or diagram model before the equation: total, difference, rate, repeated group, percentage of, or equal quantities. Then compare the operator pattern.
Signature: unknown assigned to the wrong quantity
Algebra can be executed perfectly from a badly chosen or badly interpreted variable.
Write “let x = …” with unit and meaning. If later terms do not match that meaning, the variable definition is the first weak link.
Signature: copied constant changes midway
A multi-step solution may suddenly use 36 where the stem gave 63, or 0.25 becomes 25.
Compare constants at each transition rather than re-solving the method. Copy errors have a local signature and local repair.
Signature: sign flips after moving a term
The equation setup may be correct and only one rearrangement step fails.
Mark the exact transition and practise equation balance rather than reteaching the whole topic. This is a classic execution signature.
Signature: brackets disappear too early
If only one term is multiplied or negated when the entire bracket should be affected, later coefficients show a distinctive mismatch.
Re-expand the bracket locally and compare term by term.
Signature: factorisation nearly works but middle term is wrong
The product of end terms may be correct while the sum condition fails.
Check factor pairs against both product and required middle coefficient. The error signature points to pair selection, not expansion knowledge.
Signature: substitution produces the correct magnitude but wrong sign
The formula may contain a negative coefficient, directed quantity or bracketed expression that was entered incorrectly.
Write the substituted line explicitly before calculator entry. Visible signs are easier to inspect than a hidden calculator string.
Signature: answer matches one intermediate quantity, not the final target
Multi-step questions often ask for a final cost, total distance or percentage after an earlier value is found.
Return to the exact question after the calculation. A correct intermediate result is not automatically the requested answer.
Signature: answer ignores an integer constraint
A decimal number of people, vehicles or complete packs signals that interpretation is unfinished.
Use the context to decide minimum, maximum, floor, ceiling or another whole-number action. Do not apply ordinary rounding blindly.
Signature: answer violates an ‘at least’ or ‘at most’ condition
A boundary word can reverse which integer is acceptable.
Test the nearest candidates against the original condition. The condition, not the decimal digit, decides.
Signature: rounding changes a later answer noticeably
This points to premature rounding in an intermediate step.
Return to the last fuller value or exact form and continue from there. Use the discipline from Vol 0051: preserve first, present last.
Signature: final rounded form has the wrong number of significant figures
The underlying Mathematics may be correct.
Identify the first non-zero digit and count the required significant digits. This is a presentation error unless earlier rounding contaminated the working.
Signature: angle presented to the wrong default accuracy
K210’s current notes distinguish the default presentation of angles in degrees from other non-exact answers unless the question specifies otherwise.
Read the final-answer rule before changing the calculation. The diagnostic question is presentation or process?
Signature: answer exactly matches a distractor-like shortcut
In practice materials, a wrong answer may equal what you get by omitting one step, using diameter instead of radius, or forgetting a conversion.
Work backwards from the wrong value: what single shortcut would produce it? This reverse-forensic method often exposes the missing operation quickly.
Signature: geometry result violates the diagram’s rough scale
Diagrams are not always drawn to scale, but gross impossibility still matters: a tiny segment should not become fifty times the whole under ordinary conditions.
Use geometric relationships, not visual measurement, but let rough scale trigger a check when the result is absurd.
Signature: Pythagoras uses the wrong side as hypotenuse
The resulting length may be impossible, such as a supposed hypotenuse shorter than a leg.
Identify the side opposite the right angle before writing the equation. The inequality relationship gives a fast diagnostic.
Signature: angle sum exceeds or falls short of the figure’s rule
Triangle, straight-line or polygon angle relationships leave recognizable totals.
Check the governing total before tracing arithmetic. A wrong total points to relation choice; a wrong component points to execution.
Signature: similar-shape length factor is applied to area directly
The final area may be off by a squared factor.
Ask whether the quantity is one-dimensional, two-dimensional or three-dimensional. Scale factors transform differently across length, area and volume.
Signature: statistics conclusion overstates the data
A correct mean or median can be followed by an unsupported claim such as “Group A is better in every way.”
Separate calculation from interpretation. State only what the statistic and available data justify.
Signature: probability cases are incomplete
A total below 1 when all outcomes should have been exhausted may indicate missing cases.
Build a systematic list or table and ask whether every possible category has been represented exactly once.
Signature: probability cases are duplicated
A total above 1 or an unexpectedly high favourable count can signal double counting.
Use mutually exclusive case labels or a tree structure to expose overlap.
Signature: calculator value is plausible but estimate disagrees
The error may be a hidden input mistake rather than a conceptual failure.
Compare the order of magnitude first. If estimate and calculator differ strongly, inspect entry, units and brackets.
Signature: answer changes dramatically with a tiny algebra edit
This can indicate the model is sensitive to a denominator, exponent or sign.
Do not assume the later answer is safer because it looks cleaner. Return to the original relationship and verify the sensitive term.
Signature: two methods give different answers
Method disagreement is powerful evidence that at least one route is wrong.
Compare the earliest point where the methods encode the problem differently. Do not simply average or choose the nicer answer.
Signature: two methods agree but both use the same bad assumption
Agreement is not proof if both routes inherit the same wrong base, unit or interpretation.
Use an independent check that does not share the assumption: estimate, context boundary, substitution into the original, or alternate representation.
Signature: blank after a long start
The learner may have recognised the topic but failed to identify a workable subgoal.
Write the target quantity, list known quantities, and ask what intermediate value would connect them. The signature is planning failure, not necessarily knowledge failure.
Signature: many crossed-out starts
Repeated method switching suggests recognition uncertainty.
Pause and classify the problem before calculating. Use contrast-pair practice rather than more speed drills.
Signature: correct answer only after a teacher names the method
This is a recognition gap even if the subsequent Mathematics is flawless.
Train mixed unlabeled sets where the learner must select the method from features of the problem.
Signature: one error contaminates all later parts
Multi-part dependence can make a local mistake look like a broad failure.
Mark the first wrong step and distinguish follow-through reasoning from independent later errors. Repair the source before drilling every downstream result.
Signature: later part is wrong even using the earlier result correctly
This suggests a second independent error after the first.
Do not stop diagnosis at the first slip. Check whether later reasoning remains valid given the learner’s own earlier value.
Signature: method is sound but explanation is too thin
K210 assesses reasoning and communication as well as technique.
If a question requires justification, add the mathematical reason rather than more arithmetic. The signature is communication depth, not computation.
Signature: correct result with invisible method
A bare final answer can be risky when essential working matters.
Show the relationship and key steps. The official K210 note makes working part of examination performance, not merely private scratch.
Signature: final unit conflicts with target quantity
A result in seconds when the question asks for speed, or square centimetres when it asks for length, is a strong type mismatch.
Use units as a type system. The final unit can point back to the wrong operation before you inspect numbers.
Signature: answer label names the wrong quantity
A correct number can be attached to “discount” when it is actually “sale price”.
Return to the noun in the question and label the final value explicitly. Interpretation errors often survive because the number itself looks reasonable.
Signature: solution is correct only for the drawn example
A method that works on one convenient case may fail generally.
Test a boundary or contrasting case. This connects to Vol 0043.
Signature: answer is reasonable but unsupported by working
In modelling or real-world questions, a plausible guess is not the same as a justified solution.
Rebuild the representation and show how the quantities connect. Plausibility is a check, not the method.
Signature: answer is mathematically valid but contextually irrelevant
A problem can generate several mathematically interesting quantities while asking for only one.
Return to the target sentence. The error is goal drift.
Signature: repeated errors cluster at the same transition
If signs, units, rounding or interpretation fail repeatedly in different topics, the transition itself is the training target.
Build a micro-drill around that transition and retest it in new contexts.
Signature: errors move when topic labels disappear
Strong performance on chapter worksheets but weak mixed-set performance suggests method recognition rather than procedure knowledge.
Use interleaved questions and require a one-line method choice before calculation.
Signature: errors appear only under time
If untimed work is accurate, the first failure may be control: rushed reading, calculator entry, omitted working or poor checking allocation.
Use Vol 0053 to target review rather than reteaching secure content.
Signature: errors disappear after one cue
A tiny cue that unlocks the whole solution indicates the knowledge is present but access is fragile.
Train retrieval and recognition with faded prompts until the learner can generate the cue independently.
Signature: errors survive explanation but disappear with worked examples
The learner may recognise a surface pattern without understanding the underlying relationship.
Ask for explanation, representation and transfer to a changed case before declaring mastery.
Signature: learner cannot explain why the corrected method works
A corrected answer copied from feedback is not yet a repair.
Require the learner to name the original failure, the corrected principle and a future warning sign.
Signature: same final error has different causes
Two wrong answers of 24 can arise from a bad formula, a copied value or arithmetic.
Never diagnose from the number alone. Use the visible route and question context to confirm the hypothesised cause.
Signature: different final errors share one cause
A wrong percentage, rate and graph interpretation may all come from confusing numerator and denominator roles.
Look for the common structural error across topics. High-value repair targets often cross chapter boundaries.
Build an error-signature ledger
Record the visible signature, first wrong step, error family and one future detector.
Examples: “×100 too large → unit conversion → execution → check scale before finalising” or “probability >1 → double counting → modelling → sum all cases.”
Keep the ledger short
Do not archive every wrong question forever.
Retain repeated signatures and high-value patterns. The ledger should make future diagnosis faster, not become another workbook.
Retest the signature in a new topic
After repairing a sign error in algebra, test sign control in graphs or directed change. After repairing unit inversion in speed, test unit price or density.
Transfer proves that the warning sign became portable.
Use error signatures during checking
When an answer looks suspicious, name its signature before re-solving.
“Too large by about 100”, “wrong sign”, “wrong type”, “violates boundary” or “unit mismatch” points to a targeted check and saves time.
Use error signatures after checking
If the correction works, record why. If it does not, revise the diagnosis rather than forcing the first explanation.
Error forensics is hypothesis testing.
Do not label every error careless
Careless is not diagnostic. It can hide reading, recognition, execution, interpretation or control failures.
Name the exact transition. A named error can be trained; a personality label cannot.
Do not label every error weak topic
One local sign slip does not justify reteaching an entire algebra chapter.
Use evidence across several questions before escalating the repair.
Do not assume a familiar signature always has the same cause
An answer ten times too large often suggests scale, but it could also come from a formula coefficient or copied digit.
Use the signature to narrow the search, then confirm against working.
Do not erase the wrong path during review
In training, preserve the original attempt long enough to study the failure.
A cleaned-up solution teaches less than a visible before-and-after comparison.
Use one-line postmortems
After correction, write: “First wrong step: ___. Next time I will check ___.”
The postmortem should be short enough to repeat consistently.
Use worked examples backwards
Given a wrong answer, ask the learner to invent a plausible error that would produce it.
This builds diagnostic imagination and makes common traps easier to recognise during real work.
Use contrast pairs
Place two questions that differ by one structural feature: percentage base, rate direction, graph scale, integer constraint or unit power.
Ask why the same method cannot be copied unchanged. Contrast sharpens recognition.
Use deliberate-error drills
Provide a correct solution containing one inserted error. The learner identifies the signature before locating the line.
This separates diagnostic skill from the pressure of solving the whole problem.
Use partial solutions
Show a solution up to a particular step and ask whether the next move is valid.
This trains transition checking, where many error signatures originate.
Use answer-only forensics sparingly
Sometimes give only the problem and a wrong final answer, then ask for likely causes.
The exercise builds hypotheses, but complete diagnosis still requires working. Make that limitation explicit.
Use full-paper forensics
After a timed K210 paper, cluster wrong answers by signature rather than chapter.
You may discover that units, interpretation or recognition cause more losses than any single topic.
Turn clusters into the next study plan
If four chapters all show the same copied-value or sign signature, the next week should include a cross-topic control drill.
Study plans improve when they target the first repeated failure rather than the loudest chapter label.
False-positive signatures: suspicious does not mean wrong
A signature is a clue, not a verdict. A very large answer may be correct if the units make the scale reasonable; a negative value may be valid for a directed quantity; a probability-looking decimal may actually be a rate rather than a probability. Diagnosis must return to the question before changing the answer.
Train this explicitly by including some unusual but correct answers in forensic drills. The learner should learn to ask, “What rule makes this impossible?” rather than “Does this look strange?”
Cross-topic retesting: prove the detector travels
After learning one detector, test it in at least three settings. A sign detector can appear in algebra, gradients and directed change. A unit detector can appear in speed, area, volume and density. A boundary detector can appear in probability, integer feasibility and percentage contexts.
The repair is mature when the learner recognises the signature before a teacher names the chapter.
A five-step error-forensic routine
- Describe the wrong answer’s signature: magnitude, direction, type, structure or presentation.
- Inspect the working for the earliest step capable of producing that signature.
- Run the cheapest independent check: estimate, unit, substitution, boundary or alternate representation.
- Correct the step and continue without rebuilding sound work unnecessarily.
- Record one future detector if the pattern is repeatable.
The routine is faster than indiscriminate re-solving once the learner has practised the signatures.
A worked example: answer 100 times too large
A learner calculates an area and obtains 240,000 cm² where the rough dimensions suggest about 2,400 cm². The signature is multiplicative scale. Before checking the area formula, inspect the unit conversion. If one length in metres was converted as though the area itself needed only a single factor of 100, the scale error becomes visible immediately.
A worked example: probability 1.2
The signature is boundary violation. Probabilities cannot exceed 1. Inspect whether overlapping cases were added twice or whether the denominator represents the correct total. The impossibility narrows the search before any detailed recomputation.
A worked example: correct algebra, wrong final answer
A learner solves for the discount amount correctly but the question asks for the final sale price. The signature is target mismatch. No algebra repair is needed. The learner must return to the task and complete one final interpretation step.
Use the Mathematics Hub for content repair
When the signature points to a real concept gap, return to the Mathematics Hub. Error forensics tells you where to look; it does not replace learning the underlying algebra, geometry, statistics, probability or number relationships.
The PSLE bridge
The PSLE Learner’s Guide series develops representation, unit control, checking and recovery. At G2, the larger range of algebraic, graphical and statistical work creates more distinctive error signatures, but the principle remains: read the evidence before choosing the repair.
The Examination Craft bridge
The Examination Craft hub develops checking and time allocation. Error signatures make checking cheaper because the learner can inspect the most likely fragile transition instead of restarting every uncertain question.
Readiness criteria
- You can describe how a wrong answer is wrong before re-solving.
- You use scale, sign, unit, boundary and type checks automatically.
- You locate the first wrong step rather than patching only the final line.
- You distinguish local execution errors from modelling and interpretation errors.
- You keep repeated signatures in a short error ledger.
- You retest repairs in different topics.
- You use error signatures to make final checking faster under time.
Official-source discipline
For current K210 structural and presentation rules, use the official 2027 G2 Mathematics K210 syllabus. If SEAB updates the syllabus, the current official document takes priority over this guide.
Final rule: the wrong answer is evidence
A wrong Mathematics answer is not only a failure state. It has a shape. That shape can tell you where to look, what to test and how much of the solution actually needs repair.
Read the signature before you erase the trail. Diagnose the first broken transition, fix it, and keep the warning sign for the next question. That is how mistakes become a faster checking system instead of a repeated surprise.
