The A-Math Error-Control Protocol | 10 Rules for Stopping Repeat Mistakes
A correction is not necessarily a repair.
If the student understands the teacher’s correction today and repeats the same mistake next week, the system returned to the same failure state.
The purpose of error correction is not to make the page correct. It is to make the learner less likely to recreate the error.
Rule 1: Find the First Wrong State
Ignore the red cross at the bottom for a moment. Trace the working upward and identify the first line where the mathematics became invalid.
Rule 2: Name the Mechanism
Do not label everything careless. Decide whether the failure is concept, algebra, route, execution, retrieval, transfer, condition, load or examination conversion.
Use the A-Math Error Taxonomy.
Rule 3: Repair Upstream When Necessary
If a differentiation question fails because fractions are unstable, more differentiation questions may reinforce the wrong problem. Repair the prerequisite first.
Rule 4: Correct Only What Needs Correction
Do not rebuild an entire chapter when one dependency is weak. Preserve what already works and target the smallest justified repair.
Rule 5: Close the Solution and Redo
Once the correction is understood, remove the worked answer and reconstruct the route independently. This distinguishes recognition from capability.
Rule 6: Return After Time Has Passed
A repair that works only while the correction remains in working memory is temporary. Return later and test whether the learner can rebuild it.
Rule 7: Change the Surface
Use a nearby question with changed notation, representation or context. If the repair survives only the original question, transfer is still weak.
Read the Musical Chair Syndrome.
Rule 8: Reintroduce Load
After isolated repair, place the capability back inside mixed questions, longer working and timed sets. The repair must survive realistic interference.
Rule 9: Track Recurrence, Not Just Scores
An error log should answer: What keeps returning? Under what conditions? How often? What repair has already been tried? Did it survive delayed testing?
Rule 10: Escalate Only When the Same Mechanism Survives Repair
If a targeted repair fails repeatedly, widen the diagnostic search. The visible error may be downstream of a deeper dependency, load problem or examination-state problem.
The Complete Error-Control Loop
Attempt → Detect → Locate → Classify → Repair → Redo → Delay → Transfer → Add load → Retest
That loop changes the meaning of a mistake. It stops being merely evidence of failure and becomes telemetry about the learner’s current mathematical system.
For Secondary 3 and Secondary 4
In Secondary 3, the priority is preventing weak habits from hardening. Read Secondary 3 A-Math Error Prevention.
In Secondary 4, the priority shifts toward protecting marks under examination pressure. Read Secondary 4 A-Math Mark Leakage.
For the complete system, continue to How Additional Mathematics Fails and the Additional Mathematics Master Gateway.
