BTT Mathematics / Primary Mathematics Learning Hub / Working Mathematically

Primary Mathematics: Critiquing Mathematical Arguments, Error Analysis and Comparing Reasoning | Worked Learning Guide

Critiquing mathematics means testing the reasoning, not judging the person. A strong critique identifies the exact step, assumption or interpretation that fails—and explains what would repair it.

Primary learners are often asked to explain their own thinking. Another powerful capability is evaluating someone else’s. Is the claim true? Does the evidence support it? Does the diagram actually represent the conditions? Is every case included? Has the learner proved a minimum, or merely found one good example? Is the final answer correct for the wrong reason?

Critique turns mathematical communication into a two-way process. The learner must read an argument carefully enough to reconstruct what it claims, test whether each step follows, distinguish arithmetic slips from model errors, and compare alternative methods on validity, clarity and efficiency.

The goal is not adversarial debate. It is disciplined reasoning: identify what is sound, isolate what is weak, and improve the mathematics.

Identify the claim · Find the first error · Test assumptions · Compare methods · Judge evidence · 24 questions · Worked solutions

1. Before critiquing, state what the argument claims

Argument: “3/4+1/8=4/12 because I added the numerators and denominators.”

Claim: the sum of the two fractions is4/12.

Method claimed: add numerator to numerator and denominator to denominator.

A useful critique must address both the result and the method.

Worked critique

The method combines unlike fractional units. Quarters and eighths are different-sized parts, so they cannot be counted together until rewritten in a common unit. Convert3/4=6/8, then6/8+1/8=7/8.

2. Find the first mathematical break

A later line can be wrong because of an earlier mistake.

Example:

0.7<0.65 because7<65.

First error: treating the decimal digits7 and65 as whole-number values instead of comparing tenths and hundredths.

Repair:0.7=0.70, and0.70>0.65.

The problem is place-value reasoning, not comparison-symbol knowledge.

3. Distinguish arithmetic error from model error

A learner solves: “53 pupils need vans holding8 each.”

53÷8=6 remainder5 is arithmetically correct.

Answer “6 vans” is contextually wrong.

The model interpretation failed: five pupils remain untransported, so a seventh van is required.

Why the distinction matters

More division practice would not necessarily repair this error. The learner needs remainder interpretation in discrete contexts.

4. A correct answer can come from invalid reasoning

Equation x+17=45.

Learner writes:

x=45−17=28.

This can be valid if understood as subtracting17 from both sides.

Another learner writes an invalid equality chain but happens to end at28. Substitution verifies28 as a solution, but does not make the invalid intermediate steps mathematically acceptable.

Critique the method separately from the candidate answer.

5. Check whether each equals sign is true

18+7=25×2=50−4=46 is false as one equality chain because25≠50.

Repair:

18+7=25

25×2=50

50−4=46.

Every equals sign must connect expressions of equal value.

6. Identify hidden assumptions

Claim: “The bigger rectangle must have the bigger area.”

Question: bigger in what sense? Longer? Wider? Larger perimeter? Visually larger on the page?

A critique should expose ambiguous assumptions before accepting the conclusion.

Worked example: graph

“Graph A rises more steeply on the page, so its numerical rate is larger.”

Hidden assumption: both graphs use the same axis scales.

If scales differ, visual steepness cannot be compared directly.

7. Check whether the diagram is being treated as to scale

A triangle appears isosceles but has no equal-side marks or stated condition.

Critique: appearance alone is insufficient. Mathematical properties must come from labels, stated conditions or valid deductions.

8. Test universal claims with counterexamples

Claim: “Every even number is divisible by4.”

Counterexample:6 is even and not divisible by4.

One valid counterexample refutes the word every.

The critique can then offer a repaired statement: every number divisible by4 is even.

9. Test reverse reasoning separately

True: if a number is a multiple of6, it is even.

False reverse: if a number is even, it is a multiple of6.

Counterexample8.

A critique should detect when a learner has reversed an implication without justification.

10. Check whether a finite list is complete

A learner lists whole-number factor pairs of24:

1×24,2×12,3×8.

Missing4×6.

Critique: the list has no stopping rule. Continue factors in increasing order until the pair reverses or meets.

11. Check whether cases overlap

Choosing two letters from A,B,C, learner lists AB,BA,AC,CA,BC,CB.

If order is irrelevant, each pair appears twice.

Critique: define one canonical order, such as alphabetical, to prevent duplicates.

12. Judge whether evidence is strong enough for the claim

Claim: “The sequence will always continue by adding4 because the first five terms do.”

The five terms support a pattern hypothesis but do not logically force the continuation unless the generation rule or structure is known.

A critique can distinguish consistent with evidence from proved for all cases.

13. Examples can support without proving

3+5=8,7+9=16,11+13=24 support the conjecture odd+odd=even.

A general argument is stronger: each odd number is pairs plus one; the two leftovers form another pair.

Critique asks whether the evidence matches the strength of the conclusion.

14. A proof of existence is not a proof of optimality

A learner finds a 13-move path through a grid and says “13 is the minimum.”

Finding a 13-move route proves13 is achievable.

To prove minimum, also show fewer than13 moves are impossible—perhaps because coordinate displacement already requires8 horizontal+5 vertical moves.

Critique identifies the missing lower bound.

15. A successful example is not a complete strategy

A weighing puzzle method works when the first comparison balances but fails when one pan is heavier.

Critique: a guaranteed strategy must specify and solve every possible branch, not only the convenient result.

16. Check units as part of the argument

Learner writes area=24 cm.

Arithmetic may be correct, but unit is wrong. Area requires square units:24 cm².

Units are part of mathematical meaning, not decoration.

17. Check the percentage base

Learner says “The price increased by20%, so I calculated20% of the new price.”

Critique: unless stated otherwise, a percentage increase is calculated from the original base. The new amount is not the reference whole for the increase.

18. Check discrete constraints

A calculation produces6.625 vans.

Critique: the quotient may describe capacity ratio, but vans are whole objects. Context requires interpretation and usually rounding up when all passengers must fit.

19. Compare two correct methods

25×36.

Method A: written multiplication.

Method B:100×36÷4=900.

Both valid. Method B is shorter here because25 is one quarter of100.

A critique of methods can discuss efficiency without declaring a correct longer method “wrong”.

20. Compare transparency

For A:B=3:5 total64:

Method A: bar model with eight equal units.

Method B: A=3u,B=5u,8u=64.

Both expose the same structure. The bar may be more accessible; the symbolic method more compact.

Method comparison can be about communication and cognitive load, not only speed.

21. Compare checking power

Digit-sum remainder check can reject many arithmetic errors but may miss digit transpositions.

An exact inverse operation can provide stronger verification for a specific calculation.

Critique asks what a checking method can and cannot establish.

22. Critique respectfully and specifically

Instead of “Your answer makes no sense,” write:

“The division53÷8 is correct, but six vans hold only48 pupils. The remainder represents five pupils still needing a place, so the final interpretation needs one more van.”

The critique names what is correct, where the reasoning changes, and how to repair it.

23. Ask questions that improve an argument

• What does this number represent?
• Which condition have you used here?
• How do you know your list is complete?
• Could there be a counterexample?
• Does this step preserve equality?
• What happens in the other branch?
• What does your unit mean?
• Can you check this with a different method?

A good critique often begins as a precise question rather than a correction.

24. Error analysis should end with a transferable repair

Error: 0.7<0.65.

Repair principle: align decimal place value or rename with trailing zeros.

Delayed transfer: compare0.8 and0.79.

The goal is not only to fix the original answer but to build a rule that travels.

25. Common critique errors

• Focusing on tone or handwriting instead of mathematical content.
• Calling an answer wrong without locating the first invalid step.
• Assuming a different method is invalid because it is unfamiliar.
• Using one successful example to defend a universal argument.
• Finding a counterexample that violates the claim’s conditions.
• Rejecting a correct method merely because it is longer.
• Correcting arithmetic while leaving the model error intact.
• Critiquing a learner rather than the reasoning.

26. An argument-critique protocol

1. Claim: What is being asserted?
2. Evidence: What facts, examples or steps support it?
3. Logic: Does each step follow?
4. Assumptions: What conditions are being used or hidden?
5. Coverage: Are all cases or branches handled?
6. Check: Can a counterexample, inverse method or boundary test challenge it?
7. Repair: What is the smallest change that makes the reasoning valid?

27. Practice: 24 original critique tasks

Questions 1–8: Locate the first error

1. Critique3/4+1/8=4/12.
2. Critique0.7<0.65 because7<65.
3. A learner calculates53÷8=6 r5 and answers6 vans. Critique.
4. Critique the equality chain18+7=25×2=50.
5. A rectangle area is written24 cm. Critique.
6. A graph appears steeper but uses a different scale. Critique the claim that its rate is larger.
7. A triangle looks isosceles but has no stated equal sides. Critique.
8. A percentage increase is calculated from the new value. Critique.

Questions 9–16: Test arguments

9. Disprove “every even number is divisible by4”.
10. Test the reverse of “multiple of6 implies even”.
11. A factor-pair list for24 stops at3×8. What is missing?
12. A learner lists AB and BA as different unordered pairs. Critique.
13. Five sequence terms fit “add4”. Has the continuation been proved? Explain.
14. A learner finds a13-move route and declares it minimal. What additional argument is needed?
15. A strategy handles only the balanced branch of a weighing puzzle. Is it complete?
16. Three odd-number examples have even sums. What stronger argument proves the general claim?

Questions 17–24: Compare and repair

17. Compare written multiplication and quarter-of-100 reasoning for25×36.
18. Compare a bar model and u-variable method for ratio3:5 total64.
19. Explain why a remainder-nine check passing does not prove an arithmetic answer correct.
20. Rewrite “Your answer makes no sense” as a precise critique of6 vans for53 pupils at8 per van.
21. Write a critique question for an incomplete finite list.
22. Write a critique question for a possible hidden assumption in a geometry diagram.
23. For error0.7<0.65, give one transferable repair and one delayed test.
24. Create a flawed Primary Mathematics argument and write a respectful critique identifying the first invalid step.

28. Worked solutions

1. Quarters and eighths are unlike units; rewrite3/4 as6/8, giving7/8. 2. Compare decimal place value:0.70>0.65. 3. Six vans hold48, leaving5 pupils;7 vans needed. 4. 25 does not equal50; separate the sequential equations.

5. Area unit must be cm². 6. Visual steepness cannot be compared without common axis scales. 7. Appearance alone does not establish equal side lengths. 8. A percentage increase normally uses the original base unless another reference is stated.

9. 6 is even and not divisible by4. 10. Reverse false;8 is even but not multiple of6. 11. 4×6. 12. If order is irrelevant, AB and BA duplicate one selection.

13. No. The displayed terms support but do not force the future rule without a generating structure. 14. A lower bound showing fewer than13 moves impossible. 15. No; every possible first result branch must be solved. 16. Represent each odd as pairs plus one; the two leftovers form a pair.

17. Both correct; quarter-of-100 is efficient because25=100/4. 18. Both represent eight equal units; bar is visual, u-method compact. 19. Different numbers can share the same remainder-nine value; some wrong answers pass. 20. “The division is correct, but six vans seat only48 pupils; the remaining five require a seventh van.”

21. “What stopping rule shows there are no other allowed cases?” 22. “Which stated property proves these two sides/angles are equal?” 23. Rename trailing zeros; delayed test compare0.8 and0.79. 24. Answers vary; critique should target reasoning precisely and propose a repair.

29. Critique laboratory: three answers, three different diagnoses

Problem: Find3/4 of40.

Learner A: 40÷4=10;10×3=30. Correct model and arithmetic.

Learner B: 40÷3×4≈53.3. Model error: denominator and numerator roles reversed.

Learner C: 40÷4=10;10×3=20. Model correct, arithmetic error in10×3.

The final wrong answers are different, but the more important difference is the first weak link. B needs fraction-of-quantity structure repaired; C needs arithmetic correction without reteaching the model from the beginning.

30. Parent and tutor guide

Ask learners to critique anonymous reasoning samples before critiquing their own work. This reduces defensiveness and helps build vocabulary for mathematical strengths and weaknesses.

Require one positive identification before the repair: “This step is correct because…” Then locate the first break. This encourages accurate reading rather than hunting only for faults.

Compare multiple correct methods regularly. Critique should include recognising elegance, transparency and efficiency—not only detecting errors.

31. Mastery receipt

• I identify the claim before judging it.
• I locate the first mathematical error.
• I separate model errors from arithmetic errors.
• I test hidden assumptions and reverse implications.
• I use counterexamples correctly.
• I distinguish examples from proof and constructions from optimality proofs.
• I compare correct methods on validity, clarity and efficiency.
• I critique reasoning respectfully and propose a transferable repair.

Sources and scope

Illustrative Mathematics — Construct Viable Arguments and Critique the Reasoning of Others frames mathematical critique around understanding others’ approaches, deciding whether arguments make sense and asking useful questions.

NCTM — Engaging in the Mathematical Practices highlights making conjectures, monitoring reasoning, constructing arguments, attending to precision and checking whether answers make sense.

For Singapore subject scope, use the MOE Primary Mathematics Syllabus P1–P6, updated October 2025.

Continue the apex-practices batch

• Abstract and Quantitative Reasoning
• Mathematical Structure, Decomposition, Equivalence and Rewriting
• Regularity in Repeated Reasoning and General Methods
• BTT Primary Mathematics Learning Hub

The Quiet Return

A strong mathematical community does not protect ideas from criticism. It protects people while making the ideas answerable to evidence, logic and precision.