A-Math Error Taxonomy
A wrong answer is not a diagnosis.
Two Secondary 4 Additional Mathematics students can lose the same mark for completely different reasons. One never understood the idea. Another understood it last week but could not retrieve it. Another knew exactly what to do but chose a poor route. Another performed the correct method and lost the answer through one sign error.
Before fixing an A-Math mistake, classify what kind of failure produced it.
The Six Main Error Families
For practical tuition and revision, most A-Math mistakes can be sorted into six useful families:
- Concept — the mathematical idea is not yet understood correctly.
- Retrieval — the idea was learned but cannot be produced when needed.
- Recognition — the student knows the mathematics but does not identify what the question requires.
- Method — the student recognises the problem but chooses an inefficient, incomplete or invalid route.
- Execution — the route is sound but algebra, arithmetic, notation or transcription breaks.
- Examination — time, checking, fatigue, sequencing or pressure changes the final result.
1. Concept Error
A concept error occurs when the student’s internal mathematical model is wrong or incomplete.
For example, a student may treat an identity as though it were true only for one value, misunderstand what a stationary point represents, or apply a logarithmic rule that does not exist.
More drilling does not reliably solve this because the student may simply practise the misconception faster. The repair must return to meaning, conditions and structure.
2. Retrieval Error
A retrieval error is different. The student may genuinely understand the idea when reminded, yet be unable to produce the formula, relationship or method independently.
This is why “I know it when I see it” can be misleading. Recognition during revision is not the same as recall during an examination.
Retrieval errors need closed-book reconstruction, spaced return and repeated use without prompts.
3. Recognition Error
A recognition error happens when the student possesses the required mathematics but does not see that it belongs to this question.
Chapter practice can hide this weakness because the heading already tells the student what technique to use. Mixed practice exposes it.
The repair is not another page of identical questions. It is training the student to read beneath the surface: What is given? What is constrained? What relationship is available? What mathematical object am I looking at?
4. Method Error
A method error occurs after recognition. The student sees the mathematical family but selects a poor route.
The route may be valid but unnecessarily long. It may create difficult algebra. It may ignore a condition. Or it may commit too early to an approach that cannot complete the problem.
Method errors improve through route comparison: after solving, ask which method was shortest, safest, clearest and least likely to produce secondary mistakes.
5. Execution Error
Execution errors are the mistakes most often called “careless”.
- a negative sign disappears;
- a term is copied incorrectly;
- brackets are expanded wrongly;
- an arithmetic step fails;
- notation becomes ambiguous;
- a correct expression is entered incorrectly into the calculator.
But “careless” is still too broad. If the same execution error appears repeatedly, it is a behavioural pattern that deserves a specific repair.
6. Examination Error
Sometimes the mathematics is good enough and the paper still goes wrong.
The student spends too long on one question, rushes straightforward marks later, fails to return to a skipped part, checks low-risk work while leaving high-risk algebra untouched, or becomes less accurate as cognitive load accumulates.
These are examination-control errors. They cannot be solved only by teaching another chapter.
Why “Careless Mistake” Is Not a Useful Final Label
“Careless” describes how a mistake feels. It rarely tells us what to train.
A lost sign may come from rushing, weak notation, cognitive overload, an unstable algebra habit or poor checking. Those causes require different responses.
Replace “careless” with a trainable description.
One Question Can Contain More Than One Error
Errors can cascade.
A student may first misrecognise the problem, choose the wrong route, then make an algebraic slip while trying to force that route to work. The final wrong answer contains several failures, but the earliest important one usually deserves priority.
This is why we look for the first meaningful break rather than counting every red mark independently.
Error Frequency Matters
A one-off error and a recurring error should not receive the same response.
If a student loses a sign once in twenty papers, it may be noise. If the same sign behaviour appears every week, it is a pattern. If it appears across algebra, trigonometry and calculus, it may be an upstream execution habit affecting the whole subject.
Use an Error Log as a Map, Not a Scrapbook
An error log should not become a giant collection of wrong questions.
Record the useful information:
- question or topic;
- first failure point;
- error family;
- probable root cause;
- repair attempted;
- whether the same error returned later.
Over time, several apparently different questions may collapse into one repeated error pattern. That compression is valuable because one repair may recover marks across many topics.
The Purpose of the Taxonomy
The taxonomy is not meant to make mistakes sound complicated. It does the opposite.
Instead of responding to a low score with “study harder”, we can ask a smaller question:
Was this Concept, Retrieval, Recognition, Method, Execution or Examination?
Once the error family is known, the next action becomes much easier to choose.