Did you know that one carefully chosen number can defeat a statement that seems to work for a hundred examples? Students sometimes decide a rule is true because it worked for the first few numbers they tried. Mathematics asks for a stronger check whenever a claim contains words such as every, all or always.
A counterexample is a valid example that makes a universal claim false. To test an ‘always’ statement, identify exactly which numbers or shapes the rule covers, try a case that satisfies those conditions, and check whether the claimed result fails. One genuine counterexample disproves an ‘always’ claim; several examples that work do not prove it true.
Start by diagnosing what the child thinks a test proves
Ask a learner, ‘If the rule works for 4, 8 and 12, have we proved it works for every whole number?’ A student who replies yes may be confusing pattern spotting with proof. Another learner may know that counterexamples exist but choose a number outside the stated group. These are different difficulties: the first concerns the meaning of ‘all’; the second concerns reading the condition.
Try a five-minute diagnostic. Present ‘Every multiple of 4 is a multiple of 8.’ Ask the child to say what counts as a multiple of 4, then predict whether the statement is always true. Do not initially provide a large list. Listen for whether the child searches for a failing case or looks only for more confirming cases.
Worked example: one school club statement, one decisive number
Imagine an invented classroom notice saying, ‘Every number of stickers that can be packed in groups of four can also be packed in groups of eight with no leftovers.’ Assume each sticker is whole and each group is filled equally. The sentence makes a universal claim about positive whole-number quantities divisible by four.
Test 8 stickers: 8 ÷ 4 = 2 groups and 8 ÷ 8 = 1 group. This supports the claim for eight stickers but does not settle every quantity. Try 12 stickers instead. Twelve divides into three groups of four, because 12 ÷ 4 = 3. But 12 ÷ 8 = 1 with a remainder of 4. The statement fails for a quantity that satisfies its starting condition.
Therefore 12 is a counterexample. It belongs to the relevant group (multiples of 4), yet does not have the promised property (being a multiple of 8). The single value is enough to refute the word ‘every’. It does not imply that no multiple of 4 is also a multiple of 8: 8, 16 and 24 work. The right conclusion is ‘some, but not all’.
The two-part validity check
- Check membership: Does the proposed example satisfy every starting condition? For our statement, 12 is a positive whole number divisible by 4.
- Check failure: Does the example actually contradict the conclusion? Here, division by 8 leaves a remainder, so yes.
- State the exact conclusion: ‘The always claim is false’ is justified. ‘It never works’ is not.
- Keep the domain visible: If a question concerns whole-number counts, a decimal number is not an appropriate counterexample.
A common false move is to use 7 stickers. Seven is not a multiple of 4, so it cannot refute a statement whose condition only concerns multiples of 4. An answer can be numerically correct and still be irrelevant to the claim being tested.
Why many correct examples are still insufficient
Suppose a child checks 8, 16, 24, 32 and 40 stickers. All five are divisible by both four and eight. These examples can suggest a pattern, but the child has only sampled a special subset of the multiples of four. The missing values include 4, 12, 20 and 28. Even a long list of successes may omit the case that matters.
For a universal mathematical assertion, a proof explains why the conclusion follows for every permitted case. The true reverse relationship can be proved: every multiple of 8 is a multiple of 4, because 8 × n = 4 × (2n) for any whole number n. The algebra establishes the structure of the relationship instead of relying on examples.
Progressive practice: from a guided check to independent reasoning
Guided: test the words ‘all’ and ‘some’
Write ‘All even whole numbers are multiples of four.’ Give the learner 4, 6 and 8. Four and eight fit; six is even because 6 ÷ 2 = 3 but 6 ÷ 4 is not a whole number. Help them say: ‘Six is an even whole number, but it is not a multiple of four; therefore the rule is false.’ Change the claim to ‘Some even whole numbers are multiples of four’; now four is enough to establish the ‘some’ claim.
Supported: identify a hidden exception
Try ‘All prime numbers are odd.’ Ask the learner to recall that a prime number has exactly two positive divisors, 1 and itself. Two is prime, because its positive divisors are 1 and 2; it is also even. So 2 is the counterexample. Do not teach ‘prime means odd except one’ as a definition: primality is about factors, not parity.
Independent: compare a false claim with a provable one
Set two tasks without hints. Task A: ‘Every odd whole number leaves remainder 1 when divided by 4.’ A counterexample is 3, because 3 ÷ 4 leaves remainder 3. Task B: ‘Every multiple of 6 is even.’ This one is true: writing any multiple as 6n = 2(3n) shows divisibility by 2 for a whole number n. Ask the child why Task B cannot be disproved with a legitimate example.
For transfer, invite the learner to invent an ‘always’ claim about rectangles, fractions or school timetables and challenge another learner to seek a counterexample. Do not allow a change of definition after a failing case has been found.
Common misconceptions to repair
- ‘I tested ten numbers, so it is proved.’ Ten successful trials can support a conjecture but do not establish every possible case.
- ‘Any example that does not work is a counterexample.’ It must satisfy the premise or condition first.
- ‘One exception means the opposite is always true.’ Refuting ‘all’ does not prove ‘none’.
- ‘A counterexample can prove a rule is true.’ It can refute a universal assertion; proving a universal truth requires a general argument or exhaustive check of a genuinely finite domain.
- ‘A very large number is always a stronger example.’ The smallest clear valid exception is often easier to explain.
What parents and teachers can observe
When a learner struggles, use a short sequence: underline ‘every’, circle the condition, write ‘must satisfy’ on the left and ‘must fail’ on the right, then test a candidate. Later remove the prompts. Ask for the two reasons, not merely the number. A good sign of progress is that the student rejects an invalid example without being told.
Teachers can keep a small record of whether the learner (1) states the domain, (2) picks a valid candidate, (3) checks both sides of the claim, and (4) explains the logical conclusion. Revisit the skill in a new subject a week later. Improvement means a more reliable explanation and independent checking, not a promised examination grade.
Frequently asked questions
Can one example prove that a mathematical claim is true?
Sometimes one example proves an ‘there exists’ or ‘at least one’ claim. It does not establish a general ‘all’ claim unless the domain literally has only that one eligible case.
What if the statement uses ‘if … then’ instead of ‘always’?
Treat it as a claim that every case meeting the ‘if’ condition meets the ‘then’ conclusion. A counterexample must make the ‘if’ part true and the ‘then’ part false.
Should students always search for a counterexample before proving something?
A few diagnostic examples can quickly catch a false conjecture. If no counterexample appears, move to structure or proof rather than assuming the statement must be true.
Further reading and trusted references
For related practice, read eduKate’s guide to hidden conditions in worked Mathematics, the Education Hub, and the Secondary 1 Mathematics teaching reference. A university-level explanation of universal statements and counterexamples is available from Mathematics LibreTexts.
The lasting habit
A good learner does not treat a pleasing pattern as the end of thinking. The learner asks what the claim actually covers, tries to break it fairly, and knows the difference between a supporting example, a counterexample and a proof. That is the core of independent mathematical judgement.