Mistakes and error repair can improve learning when it changes what happens on the learner’s next attempt. Students, parents and teachers searching for mistakes and error repair, study skills, memory, revision, feedback, active recall, exam preparation and independent learning need a system that connects practice to later performance rather than simply adding more activity.
This eduKateSG workbook connects mistakes and error repair with retrieval, feedback, error correction, examples, transfer and subject learning across English, Mathematics and Science. Its opening deliberately covers high-value search language while remaining easy to read and useful for a learner arriving with an immediate problem.
The central proposition is durable capability: an intervention is useful when the learner can later retrieve the relevant knowledge, select the right relationship, execute it accurately and check the result without the original support. The exercises are original teaching material rather than official examination questions or guaranteed outcomes.
Your 50-second route
Choose the chapter nearest the current failure. Keep an initial attempt. Identify the first important gap, study or correct that relationship, and then attempt a changed example. Schedule a later return without the model. If the learner can already perform accurately in mixed work, reduce special practice and let the idea operate inside normal learning.
1. A mistake is evidence, not the lesson
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
2. Find the first divergence
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
3. Concept error or execution slip
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
4. Misconceptions
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
5. Wrong rules
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
6. Incomplete answers
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
7. Overgeneralisation
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
8. Missing conditions
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
9. Representation errors
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
10. Arithmetic slips
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
11. Vocabulary misuse
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
12. Grammar errors
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
13. Reading inference errors
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
14. Writing errors
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
15. Science model errors
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
16. Mathematics model errors
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
17. Feedback after mistakes
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
18. Self-correction
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
19. Worked corrections
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
20. Counterexamples
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
21. Compare wrong and right examples
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
22. Retrieval after correction
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
23. Spacing corrections
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
24. Interleaving repaired ideas
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
25. Transfer to new problems
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
26. Error logs
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
27. When not to use an error log
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
28. Confidence and calibration
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
29. Productive struggle
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
30. When direct teaching is needed
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
31. Small-group correction
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
32. Parent response to mistakes
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
33. Peer correction
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
34. AI and automated correction
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
35. Seven-day repair cycle
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
36. Twenty-minute repair session
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
37. Mixed transfer assessment
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
38. Teaching guide
Begin with evidence from an actual task or a clearly fictional example. Ask what the learner expected to happen and what the response shows. A useful diagnosis names the relationship, condition, representation or piece of knowledge that controls the decision. Avoid turning one local error into a global judgement about ability.
Use a contrast to expose the boundary. Place a case where the idea applies beside a near case where one consequential feature changes. Ask what remains the same and what must change. If the learner lacks the underlying knowledge, teach it directly; repeated guessing is not a substitute for instruction.
Finish with delayed transfer. Change the numbers, wording, context, representation, audience or order while preserving the principle. Record whether the learner worked independently, after a cue or after a model. Immediate correction is useful, but later unprompted performance provides stronger evidence that the learning survived.
Continue through eduKateSG
Continue with How Feedback Improves Learning, How to Study Quickly | Spaced Repetition, How Error Correction Works, and the How X Works Hub.
Final checkpoint: inspect the evidence, repair or retrieve the decisive relationship, change the example and return later without the original support. The outcome is not a completed worksheet but a better independent next attempt.