How to Perform in PSLE | Learner’s Guide Vol 0948 | English, Mathematics and Science: Do Not Confuse Repeatability With Accuracy

CHAPTER 1 OF 12 · Back to contents

1. Define repeatability within the set

A method is repeatable when repetitions under the same conditions produce closely similar results. This comparison is among the results themselves.

Build the explanation by naming the repeated observations calculations readings or interpretations and the spread among repeats as evidence of repeatability. A remembered keyword is not enough: connect the observation to the mechanism and identify what would have to change for the conclusion to change.

Repair the nearest misconception by checking the accepted reference validated answer or independent standard needed for accuracy and returning to random variation systematic bias calibration rounding method agreement uncertainty and claim strength. State why the tempting alternative is too strong, incomplete or about a different stage.

• Check the repeated observations calculations readings or interpretations.
• Check the spread among repeats as evidence of repeatability.
• Check the accepted reference validated answer or independent standard needed for accuracy.
• Check random variation systematic bias calibration rounding method agreement uncertainty and claim strength.

Contents · Next

CHAPTER 2 OF 12 · Back to contents

2. Define accuracy against a reference

Accuracy concerns how close a result is to a justified target, accepted value or correctly established answer. It requires an external standard, not only agreement among repeats.

Build the explanation by naming the repeated observations calculations readings or interpretations and the spread among repeats as evidence of repeatability. A remembered keyword is not enough: connect the observation to the mechanism and identify what would have to change for the conclusion to change.

Repair the nearest misconception by checking the accepted reference validated answer or independent standard needed for accuracy and returning to random variation systematic bias calibration rounding method agreement uncertainty and claim strength. State why the tempting alternative is too strong, incomplete or about a different stage.

Contents · Previous · Next

CHAPTER 3 OF 12 · Back to contents

3. Recognise systematic error

A mis-zeroed instrument can give nearly identical readings that are all too high. Tight clustering therefore does not rule out a shared bias.

Build the explanation by naming the repeated observations calculations readings or interpretations and the spread among repeats as evidence of repeatability. A remembered keyword is not enough: connect the observation to the mechanism and identify what would have to change for the conclusion to change.

Repair the nearest misconception by checking the accepted reference validated answer or independent standard needed for accuracy and returning to random variation systematic bias calibration rounding method agreement uncertainty and claim strength. State why the tempting alternative is too strong, incomplete or about a different stage.

Contents · Previous · Next

CHAPTER 4 OF 12 · Back to contents

4. Recognise random variation

Readings scattered around a reference may have a reasonable average but weak repeatability. Both spread and location matter, and one statistic cannot replace the other.

Build the explanation by naming the repeated observations calculations readings or interpretations and the spread among repeats as evidence of repeatability. A remembered keyword is not enough: connect the observation to the mechanism and identify what would have to change for the conclusion to change.

Repair the nearest misconception by checking the accepted reference validated answer or independent standard needed for accuracy and returning to random variation systematic bias calibration rounding method agreement uncertainty and claim strength. State why the tempting alternative is too strong, incomplete or about a different stage.

Contents · Previous · Next

CHAPTER 5 OF 12 · Back to contents

5. Apply the distinction to Mathematics

Repeating the same mistaken operation produces the same wrong answer. Consistency confirms execution stability, not validity of the chosen method.

Build the explanation by naming the repeated observations calculations readings or interpretations and the spread among repeats as evidence of repeatability. A remembered keyword is not enough: connect the observation to the mechanism and identify what would have to change for the conclusion to change.

Repair the nearest misconception by checking the accepted reference validated answer or independent standard needed for accuracy and returning to random variation systematic bias calibration rounding method agreement uncertainty and claim strength. State why the tempting alternative is too strong, incomplete or about a different stage.

Contents · Previous · Next

CHAPTER 6 OF 12 · Back to contents

6. Apply the distinction to English

Several readers can repeat the same interpretation, but textual accuracy still requires words from the passage that support it. Agreement is not a substitute for evidence.

Build the explanation by naming the repeated observations calculations readings or interpretations and the spread among repeats as evidence of repeatability. A remembered keyword is not enough: connect the observation to the mechanism and identify what would have to change for the conclusion to change.

Repair the nearest misconception by checking the accepted reference validated answer or independent standard needed for accuracy and returning to random variation systematic bias calibration rounding method agreement uncertainty and claim strength. State why the tempting alternative is too strong, incomplete or about a different stage.

Contents · Previous · Next

CHAPTER 7 OF 12 · Back to contents

7. Apply the distinction to Science

Repeated measurements should use controlled conditions, while accuracy may require calibration, a known standard or a trusted comparison method. Record units and resolution.

Build the explanation by naming the repeated observations calculations readings or interpretations and the spread among repeats as evidence of repeatability. A remembered keyword is not enough: connect the observation to the mechanism and identify what would have to change for the conclusion to change.

Repair the nearest misconception by checking the accepted reference validated answer or independent standard needed for accuracy and returning to random variation systematic bias calibration rounding method agreement uncertainty and claim strength. State why the tempting alternative is too strong, incomplete or about a different stage.

Contents · Previous · Next

CHAPTER 8 OF 12 · Back to contents

8. Control the examination response

Describe the repeat pattern, compare with the reference separately, name possible random and systematic errors, and state only repeatable, accurate, both or neither as the evidence permits.

Build the explanation by naming the repeated observations calculations readings or interpretations and the spread among repeats as evidence of repeatability. A remembered keyword is not enough: connect the observation to the mechanism and identify what would have to change for the conclusion to change.

Repair the nearest misconception by checking the accepted reference validated answer or independent standard needed for accuracy and returning to random variation systematic bias calibration rounding method agreement uncertainty and claim strength. State why the tempting alternative is too strong, incomplete or about a different stage.

Contents · Previous · Next

CHAPTER 9 OF 12 · Back to contents

9. Worked examples

Worked case 1: A balance reads 52.0 g three times for a certified 50.0 g mass. Is it repeatable?

Controlled answer. Yes. The readings agree closely with one another.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Assess accuracy separately. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “Yes” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 2: Are those 52.0 g readings accurate?

Controlled answer. No. They are consistently 2.0 g above the reference.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Suggest a zero or calibration check. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “No” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 3: Three answers are all 36, but the correct answer is 42. What is shown?

Controlled answer. Repeatability of the same process, not accuracy. Agreement cannot repair a wrong method.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Find the first shared error. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “Repeatability of the same process, not accuracy” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 4: Readings 9.8, 10.0 and 10.2 compare with a reference 10.0. What can be said?

Controlled answer. They centre near the reference but show some spread. Accuracy and repeatability require separate descriptions.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Compare with 11.0, 11.0, 11.0. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “They centre near the reference but show some spread” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 5: Repair: the experiment is accurate because we got the same result every time.

Controlled answer. The experiment is repeatable; accuracy needs comparison with a valid reference. The original claim lacks an external standard.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Add a calibration measurement. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “The experiment is repeatable; accuracy needs comparison with a valid reference” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 6: Repair: everyone chose option B, so B must be correct.

Controlled answer. Agreement supports consistency, not correctness. The option must fit the question and evidence.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Check the decisive line. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “Agreement supports consistency, not correctness” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 7: What kind of error can repeated agreement fail to reveal?

Controlled answer. Systematic error. A shared bias shifts every result similarly.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Name an instrument example. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “Systematic error” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 8: What kind of error often increases spread among repeats?

Controlled answer. Random variation. Uncontrolled small changes affect repetitions differently.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Reduce one source of variation. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “Random variation” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 9: Can a method be accurate on average but poorly repeatable?

Controlled answer. Yes. Results may scatter on both sides of the reference so their mean is close.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Do not hide the spread. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “Yes” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 10: Can a method be both repeatable and accurate?

Controlled answer. Yes. Results can cluster closely around the valid reference.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. State both comparisons. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “Yes” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 11: Can it be neither?

Controlled answer. Yes. Results may be widely scattered and far from the reference.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Describe rather than label blindly. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “Yes” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Worked case 12: Why is redoing the same calculation sometimes insufficient?

Controlled answer. The same conceptual mistake may be repeated identically. Use a different checking route or inverse operation.

Mechanism. Identify the exact relationship, apply the relevant rule and preserve every stated condition. A surface clue may be compatible with several explanations, so the answer must show what makes this case decisive.

Misconception repair. Do not turn partial evidence into complete proof. Name the closest wrong answer, identify its missing condition and explain why the observed evidence supports the narrower conclusion.

Transfer. Cross-check independently. Solve the altered case from the mechanism rather than copying the previous wording.

Reverse construction. Begin with the controlled answer “The same conceptual mistake may be repeated identically” and invent a fresh prompt that would justify it for the same reason. Then change one decisive condition to create a near-nonexample. This reverse task checks whether the learner owns the boundary rather than merely recognising the original sentence.

Explanation standard. A complete explanation names the observation, connects it to the rule and limits the conclusion to what follows. It should remain valid if decorative details are removed. If the explanation depends on the story topic instead of the decisive relation, rebuild it.

Audit. Recheck the repeated observations calculations readings or interpretations, the spread among repeats as evidence of repeatability, the accepted reference validated answer or independent standard needed for accuracy, random variation systematic bias calibration rounding method agreement uncertainty and claim strength. Decide whether the evidence proves the conclusion or only supports a limited statement.

Contents · Previous · Next

CHAPTER 10 OF 12 · Back to contents

10. Independent practice

Practice 1: What is an independent check?

Explained answer. A check using a different route, reference or instrument where possible. It is less likely to reproduce the same hidden error.

Delayed transfer. Design one for a percentage answer. Cover the answer later and rebuild the chain.

Control question. Which exact condition does the nearest wrong answer miss? State it before checking.

Extension. Change one decisive feature while keeping the story similar. Explain why the conclusion now stays or changes.

Independent proof. Restate the answer without copying its wording. Your version must identify the evidence, the governing relationship and the boundary of the claim. Then write one fact that would be relevant but still insufficient on its own. This separates explanation from repetition.

Delayed comparison. Save this item beside the near-nonexample you created. On the next attempt, compare them before answering and point to the single feature that controls the difference. If more than one feature changed, revise the pair so the comparison stays fair.

Practice 2: How do units support accuracy?

Explained answer. They connect numbers to the intended quantity and expose conversion errors. A repeated bare number can still answer the wrong unit.

Delayed transfer. Audit the unit. Cover the answer later and rebuild the chain.

Control question. Which exact condition does the nearest wrong answer miss? State it before checking.

Extension. Change one decisive feature while keeping the story similar. Explain why the conclusion now stays or changes.

Independent proof. Restate the answer without copying its wording. Your version must identify the evidence, the governing relationship and the boundary of the claim. Then write one fact that would be relevant but still insufficient on its own. This separates explanation from repetition.

Delayed comparison. Save this item beside the near-nonexample you created. On the next attempt, compare them before answering and point to the single feature that controls the difference. If more than one feature changed, revise the pair so the comparison stays fair.

Practice 3: How does rounding affect comparison?

Explained answer. Coarse rounding can make different results look identical. Inspect unrounded values and required precision.

Delayed transfer. Separate display precision from true accuracy. Cover the answer later and rebuild the chain.

Control question. Which exact condition does the nearest wrong answer miss? State it before checking.

Extension. Change one decisive feature while keeping the story similar. Explain why the conclusion now stays or changes.

Independent proof. Restate the answer without copying its wording. Your version must identify the evidence, the governing relationship and the boundary of the claim. Then write one fact that would be relevant but still insufficient on its own. This separates explanation from repetition.

Delayed comparison. Save this item beside the near-nonexample you created. On the next attempt, compare them before answering and point to the single feature that controls the difference. If more than one feature changed, revise the pair so the comparison stays fair.

Practice 4: Why record all repeats?

Explained answer. The spread is evidence about consistency and possible outliers. A selected favourite reading hides variation.

Delayed transfer. Use a table. Cover the answer later and rebuild the chain.

Control question. Which exact condition does the nearest wrong answer miss? State it before checking.

Extension. Change one decisive feature while keeping the story similar. Explain why the conclusion now stays or changes.

Independent proof. Restate the answer without copying its wording. Your version must identify the evidence, the governing relationship and the boundary of the claim. Then write one fact that would be relevant but still insufficient on its own. This separates explanation from repetition.

Delayed comparison. Save this item beside the near-nonexample you created. On the next attempt, compare them before answering and point to the single feature that controls the difference. If more than one feature changed, revise the pair so the comparison stays fair.

Practice 5: What is the nearest wrong shortcut?

Explained answer. Treating close agreement as proof of truth. Shared methods can share the same bias.

Delayed transfer. Seek an external standard. Cover the answer later and rebuild the chain.

Control question. Which exact condition does the nearest wrong answer miss? State it before checking.

Extension. Change one decisive feature while keeping the story similar. Explain why the conclusion now stays or changes.

Independent proof. Restate the answer without copying its wording. Your version must identify the evidence, the governing relationship and the boundary of the claim. Then write one fact that would be relevant but still insufficient on its own. This separates explanation from repetition.

Delayed comparison. Save this item beside the near-nonexample you created. On the next attempt, compare them before answering and point to the single feature that controls the difference. If more than one feature changed, revise the pair so the comparison stays fair.

Practice 6: How should a comprehension interpretation be checked?

Explained answer. Quote or paraphrase the exact textual evidence and test alternatives. Repeated assertion is not evidence.

Delayed transfer. Locate a contradicting line if present. Cover the answer later and rebuild the chain.

Control question. Which exact condition does the nearest wrong answer miss? State it before checking.

Extension. Change one decisive feature while keeping the story similar. Explain why the conclusion now stays or changes.

Independent proof. Restate the answer without copying its wording. Your version must identify the evidence, the governing relationship and the boundary of the claim. Then write one fact that would be relevant but still insufficient on its own. This separates explanation from repetition.

Delayed comparison. Save this item beside the near-nonexample you created. On the next attempt, compare them before answering and point to the single feature that controls the difference. If more than one feature changed, revise the pair so the comparison stays fair.

Practice 7: How should a conclusion be worded?

Explained answer. The results were closely repeatable under these conditions and were accurate or inaccurate relative to the stated reference. The two claims remain visible.

Delayed transfer. Avoid the vague word reliable alone. Cover the answer later and rebuild the chain.

Control question. Which exact condition does the nearest wrong answer miss? State it before checking.

Extension. Change one decisive feature while keeping the story similar. Explain why the conclusion now stays or changes.

Independent proof. Restate the answer without copying its wording. Your version must identify the evidence, the governing relationship and the boundary of the claim. Then write one fact that would be relevant but still insufficient on its own. This separates explanation from repetition.

Delayed comparison. Save this item beside the near-nonexample you created. On the next attempt, compare them before answering and point to the single feature that controls the difference. If more than one feature changed, revise the pair so the comparison stays fair.

Practice 8: State the full method.

Explained answer. Compare repeats with one another, compare their location with a valid reference, inspect random spread and systematic bias, use an independent check, and label repeatability and accuracy separately. Do not let agreement substitute for evidence.

Delayed transfer. Apply across all three subjects. Cover the answer later and rebuild the chain.

Control question. Which exact condition does the nearest wrong answer miss? State it before checking.

Extension. Change one decisive feature while keeping the story similar. Explain why the conclusion now stays or changes.

Independent proof. Restate the answer without copying its wording. Your version must identify the evidence, the governing relationship and the boundary of the claim. Then write one fact that would be relevant but still insufficient on its own. This separates explanation from repetition.

Delayed comparison. Save this item beside the near-nonexample you created. On the next attempt, compare them before answering and point to the single feature that controls the difference. If more than one feature changed, revise the pair so the comparison stays fair.

Contents · Previous · Next

CHAPTER 11 OF 12 · Back to contents

11. Delayed transfer

On day one, solve clear examples. On day three, mix examples, near-nonexamples and incomplete evidence. On day seven, solve without headings or a word bank. Delay reveals whether the method can be reconstructed independently.

Keep an error log for the first failed stage: reading, representation, rule, execution or checking. The earliest failure identifies the repair.

Interleave this skill with neighbouring ideas that share vocabulary. Classify each item only after naming the distinguishing condition.

In timed work, mark the relationship, apply the rule, write the conclusion and audit one condition. If checking changes the answer, record the overlooked feature.

Create a teach-back card with a near-nonexample on the front and the repaired mechanism on the back. Explain why the tempting answer fails.

Finish with one basic example, one boundary case and one transfer case. Supply explanations that a classmate can verify.

For a second delayed test, reverse the direction: start from a conclusion and invent evidence that would justify it, then evidence that would be insufficient. This exposes the boundary between relevance and proof.

Contents · Previous · Next

CHAPTER 12 OF 12 · Back to contents

12. Family guidance and FAQs

Guidance for parents and tutors

Ask for the decisive condition before confirming the answer. Use prompts such as “What exactly happened?”, “Which stage or meaning is this?” and “What would prove it?”

Fade support from labelled examples to mixed questions and delayed open response. Independent retrieval after time and topic change is the target.

Praise precise claim control. Narrowing an answer when evidence is incomplete is scientific and academic strength.

Frequently asked questions

Should I memorise the rule?

Memorise the compact rule, then practise examples and near-nonexamples so the conditions remain visible.

What if two answers seem possible?

Return to the exact relationship and choose the option satisfying every stated condition without adding facts.

Why delay practice?

Immediate repetition may depend on the recent explanation; delay tests reconstruction.

How should an error be corrected?

Name the first failed stage and practise it with a changed example.

When is the skill ready?

When the learner can explain, apply and transfer it after a delay.

What is the final audit?

Check relationship, mechanism, claim strength and whole-question fit.

Next routes

Continue at the English, Mathematics and Science learning hub or return to the series directory.

Contents · Previous