An error log you can use straight away
- Question and date: identify the exact independent attempt.
- First wrong step: circle the earliest error, not only the wrong final answer.
- Cause and rule: name the misunderstanding and replacement principle.
- Changed question: solve a new example without the model answer.
- Delayed retest: record whether the improvement survives several days later.
Find the repair you need
- Did you know? A pile of corrected papers is not automatically a revision plan
- Start with the first wrong line, not only the final answer
- The six error categories that make a useful log
- Worked diagnosis 1: the wrong original percentage base
- Worked diagnosis 2: the minus sign inside a bracket
- Worked diagnosis 3: a circle theorem applied without checking the arc
- Worked diagnosis 4: average speed with a hidden stop
- Worked diagnosis 5: a correct value but missing units
- An example error-log row, written in ordinary language
- Why merely copying the correction does not settle the issue
- Do not confuse a conceptual gap with an exam-time decision
- How to measure improvement without invented percentages
- A 90-minute three-pupil tutorial built around real scripts
- What parents can ask after receiving a marked paper
- A seven-day error-log reset
- Frequently asked questions about a Mathematics error log
Did you know? A pile of corrected papers is not automatically a revision plan
A student may copy a model solution neatly, agree that the correct answer makes sense and then repeat the exact same mistake on a different question two days later. The completed correction page looks impressive, but the method has not yet become independent.
Quick answer: build a one-page Mathematics error log containing the question, first wrong step, cause, replacement rule, unseen retest and later outcome. Classify what failed before assigning more practice. An error caused by reading the wrong percentage base should not receive the same repair as a lost algebraic minus sign.
The 2027 SEC G3 Mathematics K310 syllabus examines standard techniques, problem solving and reasoning. It gives approximate AO1/AO2/AO3 weightings of 45%, 40% and 15%. An error log is a teaching method for those capabilities, not an official SEAB submission or mandatory examination component.
Start with the first wrong line, not only the final answer
Suppose a student calculates a discount correctly but reports the wrong final figure. The first question is not ‘How many marks were lost?’ It is ‘At what line did the mathematics first stop following the question?’
One incorrect line can create several later wrong numbers. Correcting every downstream number without fixing the earliest misconception wastes effort. Ask the student to annotate the first unjustified operation and explain the rule that should replace it.
Did you know? Two students with the same final answer may have different reasoning quality. One may use valid steps; another may accidentally arrive at the right number by cancelling terms illegally. An error log can record reasoning quality, not just whether a numerical answer matched.
The six error categories that make a useful log
Reading and representation: misunderstood what the question asked or incorrectly translated words into a diagram or equation. Concept choice: selected a formula whose conditions do not apply.
Algebraic or arithmetic execution: lost signs, mishandled fractions, expanded brackets incorrectly, or mistyped a calculation. Mathematical communication: omitted a necessary reason, unit, condition or conclusion.
Accuracy and timing: rounded too early, used a wrong angle mode, overlooked a conversion, or left a compulsory question untouched because of pacing. These labels can overlap; record the earliest cause that is worth teaching and any secondary issues.
Worked diagnosis 1: the wrong original percentage base
A fictional jacket costs S$96 after a 20% discount. A student writes ‘original = 96 × 1.2 = 115.20’. The calculation 96 × 1.2 is arithmetically accurate, but the model is wrong.
The sale price represents 80% of the original, so the correct equation is 0.8P = 96. Therefore P = 96/0.8 = S$120. Check: S$120 − 20% of S$120 = S$96.
The log should say cause: treated the new 80% as the original 100%. Replacement rule: identify the reference amount for every percentage. New retest: an item costing S$72 after a 10% discount originally cost S$80.
Worked diagnosis 2: the minus sign inside a bracket
A learner expands −2(3x − 5) + 4x as −6x − 10 + 4x, giving −2x − 10. The first wrong step is the distribution of −2 to −5.
The correct product is (−2)(−5)=+10, so the expression becomes −6x+10+4x = −2x+10. Test x=2: the original gives −2(6−5)+8 = 6, matching −4+10=6.
The log should say cause: negative × negative treated as negative, not the vague label ‘careless’. Retest with a changed expression such as −3(2x−4)+x, which correctly simplifies to −5x+12.
Worked diagnosis 3: a circle theorem applied without checking the arc
A student sees a central angle of 124° and writes that an angle at the circumference must be 62°. That conclusion is valid only when the specified circumference angle subtends the same arc and occupies the appropriate geometric position.
If the angle does stand on the same minor arc from the opposite side of the circle, the angle at the circumference is 124/2 = 62°. But a different angle standing on the major arc needs a different interpretation.
Cause: theorem selected by visual resemblance, rather than verified chord and arc. Retest by moving the circumference point to a different part of a new circle diagram and asking the learner to name the arc before calculating.
Worked diagnosis 4: average speed with a hidden stop
A journey covers 12 km at 30 km/h, pauses for 12 minutes and then covers 18 km at 45 km/h. A student divides 30 km by 48 minutes and reports a moving-only average of 37.5 km/h when the question actually asks for the entire elapsed journey.
The two legs each take 24 minutes, so moving time is 48 minutes. Add the 12-minute stop: total elapsed time is 60 minutes. The whole-trip average is 30 km/h.
The log should say cause: omitted waiting time from the requested whole-journey interval. A changed retest could alter the wait to 18 minutes and ask explicitly for both moving-only and overall averages.
Worked diagnosis 5: a correct value but missing units
A triangle has base 7.5 cm and perpendicular height 4 cm. Its area is ½×7.5×4 = 15 cm². Writing only ’15 cm’ uses a length unit for an area.
The numerical arithmetic is correct; the mathematical answer is not fully expressed. A short replacement rule is length cm, area cm², volume cm³. Ask which dimension the formula measures before calculating.
Retest with a right circular cylinder of radius 2 cm and height 5 cm: its volume is 20π cm³, not 20π cm². The new question checks whether the dimension rule transfers.
An example error-log row, written in ordinary language
Date: after a mixed practice script. Question: reverse percentage. First wrong line: multiplied the discounted price by 1.2. Cause: wrong 100% base. Replacement: final = original × 0.8 for a 20% discount.
Immediate test: student correctly finds that S$72 after 10% off came from S$80. Later retest: two days later, a different 15%-off problem is attempted without seeing the earlier worked answer. Outcome: record whether the starting base and equation were independently correct.
This is enough detail to teach from. The log does not need a large spreadsheet with dozens of decorative columns. Six short fields can make the next revision session purposeful.
Why merely copying the correction does not settle the issue
The learner can often reproduce a correct solution while it is still visible. That demonstrates the ability to follow a model, not necessarily the ability to retrieve the relationship or choose it in a changed situation.
Use three passes: first understand the actual error; second solve a fresh problem with the same underlying principle; third return after a delay and solve another question with hints removed. If the third fails, the error remains active.
Do not turn the delayed retest into a punishment or a prediction of final grade. It is a practical way to identify what needs another explanation.
Do not confuse a conceptual gap with an exam-time decision
A student may know a theorem but spend too long trying one irrelevant calculation, leaving several later questions unattempted. That problem may involve method-selection or pacing, not complete ignorance of the chapter.
Record the time spent on a blocked question and the first decision that was unproductive. Practise identifying useful information and moving to a reachable part when appropriate under the paper instructions.
The 2027 K310 paper format includes compulsory questions in both papers. A time-management technique should therefore support attempting the full paper without sacrificing valid working on problems the student can solve.
How to measure improvement without invented percentages
Track whether a previously mistaken idea is solved unaided on a changed item, then whether the same skill survives a later mixed-topic question. Record dates and outcomes. The number of independent successes is useful even without converting it into a predicted grade.
If three recent fraction mistakes were corrected but an unfamiliar fractional equation still produces a zero denominator, the domain concept needs another repair. If a student now consistently checks restrictions, that is observable progress.
A practice score can fluctuate with paper difficulty. Explain what actually changed: fewer repeated sign errors, valid models chosen without hints, stronger diagram justifications or more compulsory questions attempted.
A 90-minute three-pupil tutorial built around real scripts
An illustrative lesson might begin with 10 minutes of independent recall, 20 minutes analysing recent scripts, 25 minutes of individual concept repair, 20 minutes of mixed unseen work and 15 minutes of explanation and setting delayed checks.
One learner’s log may show a weak percentage base, another’s may show algebraic expansion signs, and the third’s may show errors in modelling speed and time. Close feedback is useful when it leads to a specific next task for each learner.
This schedule is an educational example, not a claim that every eduKateSG lesson uses identical minute allocations. The original programme remains the reference for its actual class structure and venue.
What parents can ask after receiving a marked paper
Instead of starting only with the percentage score, ask: Which two mistakes were repeated? What was the first wrong line? What rule would prevent it next time? When will you test that rule with a different question?
The conversation should end with a manageable action, not an impossible list of every missed chapter. For example: ‘Before Sunday, solve two changed reverse-percentage problems and explain which amount is 100%.’
Students can contribute to their own log by choosing the error category and explaining a correction in their own words. That supports self-monitoring while keeping the parent informed about concrete progress.
A seven-day error-log reset
Day 1: collect the latest independent school script and identify four earliest wrong lines. Day 2: label the causes and choose two high-impact repairs. Day 3: reteach the first concept and solve one changed problem.
Day 4: practise the second concept in a different form. Day 5: combine both concepts in a mixed mini-set. Day 6: review correct work and mark any remaining instability. Day 7: perform an independent delayed retest and plan the next week’s two priorities.
Keep the work proportional to school commitments and rest. A short cycle that leads to independent mastery is more useful than collecting a larger pile of unexamined answers.
Frequently asked questions about a Mathematics error log
Should every wrong question go into the log? Focus on recurring or consequential errors first rather than documenting everything without a plan.
Is a copied correction enough? No. Follow it with a changed unseen question and a delayed independent retest.
Does a good error log predict the student’s SEC grade? No. It supports targeted learning but does not guarantee outcomes.
How is this different from simply doing more past papers? The log identifies what failed and selects the corrective practice. Past papers can provide evidence, but the correction process must be deliberate.
What if two students have the same test mark? Their first wrong steps may be different, so they may need different tutorial tasks.
Is error logging an official SEAB component? No. It is an educational revision method aligned with the syllabus’s emphasis on correct technique, problem solving and reasoning.
Original small-group Mathematics teaching reference and Bukit Timah location
The unchanged eduKateSG Mathematics tutorial reference describes premium three-pupil weekly classes of 1.5 hours, guided correction and individual practice at 8 Fourth Avenue, Singapore 268674, near Sixth Avenue MRT. This article does not assert another branch or promise that a place is available.
The same error log can travel between a student’s schoolwork, independent home practice and tutor consultation. The specific question and first wrong step give the tutor enough information to plan a short repair and subsequent unseen retest.
Enquire about an eduKate Singapore parent–student Mathematics consultation.
Official examination sources and next learning articles
- SEAB 2027 SEC G3 Mathematics K310 syllabus and assessment objectives
- SEAB 2027 G3 subjects and codes
- eduKateSG original three-pupil Mathematics programme reference
- AO1, AO2 and AO3: What the SEC Tests
- Paper 2 Excursion Planning: Worked Case Study
- G3 Mathematics vs Additional Mathematics: K310 and K341
These are original revision examples, not official examination questions or grade guarantees. SEC G3 K310 begins in 2027; actual 2026 O-Level candidates should follow the appropriate 4052 syllabus and instructions.