A repeated mistake is not simply evidence that a student was careless again. Often it is evidence that the previous correction never changed the underlying model.
Correction is not the same as repair
A common sequence is simple: a student answers incorrectly, sees the correct answer, understands it while looking at it and moves on. The exercise is marked as corrected. But nothing has established that the student can recognise the same underlying problem when it appears in a different form.
Find the cause, not only the wrong answer
Errors have different causes. Knowledge may be missing. A concept may be misunderstood. The learner may know the concept but choose the wrong procedure. A question may be misread. Working memory may be overloaded. Vocabulary may conceal the task. Timing may force a poor decision. Each cause requires a different repair.
Why repetition can hide the problem
Repeating nearly identical questions can make performance improve because the pattern becomes predictable. That does not necessarily mean the underlying knowledge has become flexible. Change the wording, representation or context and the old error may return.
Across the curriculum
A Mathematics sign error may actually be a weak algebraic model. A Science explanation may fail because the learner lists facts instead of connecting cause and effect. A comprehension answer may fail because evidence is copied without answering the question. A composition may repeatedly receive the same comment because the student has never practised the specific revision operation the comment demands.
Use an error loop
A stronger sequence is: identify the error, classify its cause, explain why the original response failed, construct the correction, attempt a varied example without help, and retest after a delay. The final two steps matter because they reveal whether the repair survives beyond the original correction.
Mistakes are information
The objective is not a classroom in which nobody makes mistakes. It is a learning system in which mistakes become high-quality evidence. When the same mistake returns, do not merely correct it again. Investigate why the previous correction failed to change the next decision.
