When the same mistake returns, adults often conclude that the child was careless, did not listen or did not practise enough. Sometimes that is true. Often it is not precise enough to be useful. A repeated error is evidence that something in the learner’s internal model, checking routine or response habit has survived the previous correction.

A correction is not the same as a repair

A teacher can mark an answer wrong and show the correct answer. A parent can explain it again. A child can nod. None of those events proves that the mechanism producing the error has changed. The real test comes when the child meets a similar problem later without the adult present.

Signs to notice

• The same mathematical sign, operation or step is repeatedly mishandled.
• Comprehension answers repeatedly miss the same relationship even across different passages.
• Writing corrections are understood during conferencing but disappear in the next composition.
• Science explanations repeatedly omit the same causal link.
• The child can correct an error when prompted but rarely detects it independently.
• Practice scores plateau even though the volume of completed work rises.

Find the error family, not just the wrong answer

Collect several examples. Ask whether they share a cause. One error may be a slip; six structurally similar errors are a pattern. Useful families include missing knowledge, misunderstood concepts, misread instructions, weak procedure, language misunderstanding, poor checking, transfer failure and performance-pressure errors.

This classification matters because different causes require different repairs. More arithmetic practice will not solve a language misunderstanding. Re-explaining a concept may not solve a child’s habit of skipping command words. Telling a child to “be careful” rarely supplies a checking procedure.

The replay test

After a mistake, ask the child to reconstruct what they were thinking before seeing the correction. The purpose is not interrogation. It is to locate the decision that made the answer reasonable to the child at that moment. Then ask what signal should alert them next time.

Turn feedback into a changed next attempt

Use a short loop: identify the error, name its cause, demonstrate the repair if necessary, have the child redo the original problem without copying, then give a slightly different problem. Return to the pattern later. Improvement is demonstrated when the child can recognise the trigger and execute the repair independently.

Why mixed examples matter

If ten consecutive questions all demand the same procedure, the worksheet itself tells the learner what to do. Real examinations are less generous. Mixed practice forces the learner to decide which knowledge or procedure applies. That decision is often where repeated mistakes become visible.

What parents can record

A tiny error log is enough: date, task, visible error, likely cause, repair attempted and whether the child succeeded later without prompting. The purpose is not to catalogue failure. It is to see whether the same mechanism crosses subjects or disappears after a targeted intervention.

What not to do

Avoid attaching character labels to error patterns. “Careless,” “lazy” and “weak” compress many possible mechanisms into one judgement. Avoid correcting everything simultaneously. A child who receives twelve unrelated corrections may remember none. Prioritise the error that most limits subsequent performance.

When to seek broader help

If persistent difficulties appear across many settings, substantially interfere with learning, are accompanied by broader concerns, or remain despite appropriate teaching and targeted support, bring the pattern to the child’s teachers and, where appropriate, qualified professionals. An error log can describe what is happening; it cannot diagnose why.

The principle

A repeated mistake is not merely another wrong answer. It is a clue. Find the mechanism that keeps reproducing it, change that mechanism, and then test whether the change survives in a new situation.