JC2 H3 Mathematics Tutor Bukit Timah

Did you know? The expression n²+n+41 gives a prime number when n=0, 1 and 2. That is striking, but does it prove the expression is always prime? No. At n=40 the expression is 40²+40+41=1681=41², which is composite. The most interesting part of advanced Mathematics often begins precisely where a persuasive pattern stops working.

A strong JC2 H3 Mathematics student in Bukit Timah needs to do more than recognise patterns: they must formulate precise conjectures, try to break them, refine their scope and turn valid results into proofs. The Singapore–Cambridge H3 Mathematics (9820) syllabus explicitly includes mathematical investigation, generalisation, special cases, and completing or critiquing mathematical solutions.

Quick answer: A conjecture is a mathematical claim suggested by evidence but not yet proved. To investigate one, state the domain, test meaningful cases, search for counterexamples, revise the claim if needed, then produce a general argument. A single valid counterexample disproves a universal conjecture; a thousand confirming examples do not prove it.

A practical investigation workflow

Start with a precise question. If you notice that a sequence behaves in a particular way, write down what you suspect using a clear domain. ‘For every integer n≥1’ is a testable mathematical statement; ‘this always seems to work’ is not.

Next, calculate a handful of small examples to understand the structure. Include boundary cases, zero and negative values where the domain permits them, and cases just beyond the first examples. Use a table or a graph as a discovery tool, not as the final justification.

Then look deliberately for a counterexample. Change a sign, an exponent, a parity condition or an equality case. If the conjecture fails, ask what was missing: perhaps the statement needs nonnegative inputs, or the result holds only for odd numbers.

If the conjecture survives reasonable checks, search for a general argument: algebraic factorisation, induction, cases, contradiction, bijection, an inequality or a symmetry. The best technique is the one whose assumptions match the problem.

Finally, communicate the result in three layers: what was observed, what was conjectured and what has actually been proved. Do not quietly promote an observation to a theorem.

This is the investigative habit explicitly expected by H3 Mathematics: mathematicians formulate claims, evaluate reasoning and respond to unfamiliar statements rather than merely reproduce procedures.

The prime-number trap: convincing evidence is still not proof

Observation: Consider f(n)=n²+n+41 for nonnegative integers n. For n=0 we obtain 41, for n=1 we obtain 43, and for n=2 we obtain 47. All three are prime.

Premature conjecture: ‘f(n) is prime for every nonnegative integer n.’ This is precise enough to test, but we have not established it.

Search for a counterexample: Try n=40. Then f(40)=1600+40+41=1681=41×41. Since 1681 has factors other than 1 and itself, it is not prime.

Conclusion: The universal conjecture is false. The example at n=40 refutes it completely, even though many smaller values may have looked persuasive.

A useful extension is to ask what happens at multiples of 41. Put n=41k. Then n²+n+41=41(41k²+k+1), so for nonnegative integer k the expression is divisible by 41. When k≥1 it exceeds 41, and is therefore composite. This is a new, proved structural observation—not merely another list of computed values.

A teacher can use this example to diagnose whether students know the logical difference between ‘it worked every time I tried’ and ‘it works for every possible input’.

An example where the pattern really can be proved

Write the first few sums of consecutive positive odd numbers: 1=1²; 1+3=4=2²; 1+3+5=9=3²; 1+3+5+7=16=4². This suggests the conjecture that the sum of the first n positive odd numbers equals n².

State it precisely: for every positive integer n, Σ from k=1 to n of (2k−1)=n². The examples generate the idea, but we still need a proof.

One proof uses arithmetic series: the sum is 2Σk−Σ1 = 2[n(n+1)/2]−n=n². This is valid for every positive integer n, so it establishes the universal claim.

Another proof uses mathematical induction. The base n=1 holds. If the sum of the first k odd terms is k², then adding the next odd term 2k+1 gives k²+2k+1=(k+1)². Hence the claim holds at k+1.

A third viewpoint uses consecutive square differences: (k+1)²−k²=2k+1. The odd numbers measure the increments between successive perfect squares. This connects the pattern to the method of differences.

Comparing the proofs is itself a mathematical investigation. They arrive at the same result with different structures—an arithmetic formula, induction and telescoping.

Extension and generalisation: enlarge the claim, not the gap

A common way to develop an H3 investigation is to move from a special case to a broader one. Suppose you prove that the sum of three consecutive integers is divisible by three. Let the integers be n,n+1,n+2. Their sum is 3n+3=3(n+1), proving the claim for every integer n.

Can you generalise? The sum of any odd number m of consecutive integers is divisible by m. To show this, write m=2r+1 and take integers n,n+1,…,n+2r. Their average is the central integer n+r, so their sum is m(n+r).

Why restrict m to be odd? If m=2, the sum n+(n+1)=2n+1 is odd and therefore not divisible by two. The naive claim ‘the sum of any m consecutive integers is divisible by m’ is false.

This gives a complete cycle: begin with three terms, propose a generalisation, find an even-m counterexample, then prove the corrected odd-m statement.

Another direction is to change the operation. Three consecutive integers have a product divisible by six because at least one is divisible by three and at least one is even. That uses different reasoning from the sum result, so it should be proved separately.

A genuine generalisation identifies exactly which feature survives and why. Adding more words to a conjecture without adding the corresponding justification is not progress.

How to critique a proposed solution

Faulty argument: ‘If a²=ab, cancel a from both sides to obtain a=b.’ That sounds plausible. But what if a=0? Then a²=ab becomes 0=0 for every real b, including b=7, yet a≠b.

Locate the first invalid step: Cancelling a divides by a. Division by zero is not allowed, and the argument never established a≠0.

Repair the conclusion: Factor a²−ab=a(a−b)=0. By the zero-product property, either a=0 or a=b. If the problem additionally assumes a≠0, then a=b follows.

The repair is more informative than simply marking the original argument wrong. It identifies the exact missing condition and produces a correct general statement.

Second faulty argument: ‘If x²=y², then x=y.’ This ignores the case x=−y. Factor x²−y²=(x−y)(x+y)=0. The correct conclusion is x=y or x=−y.

When asked to critique a proof, use four questions: Which line is unsupported? What assumption is missing? Can you construct a counterexample? What is the strongest statement that follows from the valid steps?

Boundary cases and special cases can change the entire answer

Consider the claim ‘For all x>0, x+1/x≥2.’ It is true by AM–GM. If the condition x>0 is replaced with x≠0, the claim becomes false: x=−1 gives −2, not a number at least two.

Now consider ‘Every integer n has a positive square n².’ The correct statement is n²≥0 for all integers, and n²>0 only for n≠0. Testing the boundary value n=0 distinguishes a strict inequality from a non-strict one.

Or examine a theorem with a denominator x−1. If x=1 belongs to the claimed domain, the expression is undefined. An apparently trivial excluded value can determine whether the entire statement is meaningful.

Special cases are not merely afterthoughts. Sometimes they suggest a proof by cases, while other times they expose the exact hypothesis needed for a theorem.

One useful exam habit is to check n=0, n=1, negative values, equality cases and any point where a denominator vanishes—when the statement’s domain allows those inputs. Choose cases because they stress the structure, not because they happen to be easy.

What counts as evidence, and what counts as proof?

A numerical table can reveal a pattern. A graph can make an invariant visible. A computer search can discover small counterexamples. Each is useful evidence, but the final status of the claim depends on the logical argument.

Suppose a student checks a divisibility claim for the first million positive integers. That is impressive verification, yet it does not establish a universal claim over infinitely many integers. A general algebraic identity or induction proof can settle all cases in a few lines.

Conversely, when a claim is false, a single counterexample within the permitted domain is enough. More counterexamples can explain why it fails, but are not needed to reject it logically.

An investigation often benefits from a deliberate order: explore, conjecture, attack, prove or refute, then refine. That order helps avoid writing a beautiful proof of a statement different from the one actually asked.

The official H3 syllabus includes forming conjectures, extensions, generalisation, special cases and completing or critiquing a solution. These are activities students should practise directly, not simply read about.

Twelve original investigation clinics

1. Sum of odd integers

1+3+⋯+(2n−1)=n² for positive integers n. The result follows by arithmetic-series algebra or induction.

2. An attractive false pattern

n²+n+41 is prime at n=0,1,2 but not at n=40, where it equals 41².

3. Three consecutive integers

n+(n+1)+(n+2)=3(n+1), so the sum is divisible by three for every integer n.

4. Even-length counterexample

Two consecutive integers sum to 2n+1, which is never divisible by two.

5. Odd-length generalisation

An odd number m=2r+1 of consecutive integers has sum m(n+r), divisible by m.

6. Product of three consecutive integers

(n−1)n(n+1) is divisible by six because it has an even factor and a factor of three.

7. Faulty cancellation

From a²=ab one may conclude a=0 or a=b, not necessarily a=b.

8. Missing square root sign

From x²=y² conclude x=y or x=−y, not just x=y.

9. Missing zero

For every real x, x²≥0. The stronger claim x²>0 is false at x=0.

10. Missing positivity

x+1/x≥2 for x>0, but not for all nonzero real x; x=−1 is a counterexample.

11. Cycle of last digits

Positive powers of two have last digits 2,4,8,6 repeating. A proof uses 2^(n+4)−2ⁿ=15·2ⁿ, divisible by ten for n≥1.

12. Equality deserves testing

If AM–GM gives a lower bound, solve the equality condition and verify it lies within the problem’s constraints before calling the bound a minimum.

Five errors that turn exploration into a weak answer

1. Stating a pattern without a domain. ‘For all n’ leaves the reader guessing whether n is integer, natural or real. 2. Equating examples with a proof. Finite checks do not settle an infinite universal claim.

3. Failing to attack the conjecture. A mathematical investigation should look for counterexamples and special cases, not only gather supportive data. 4. Overgeneralising from a familiar condition. An inequality for positive numbers may fail for negative numbers.

5. Criticising the last line instead of the first wrong one. In a flawed proof, the most useful diagnosis is the earliest unjustified inference. Repairing it may require changing the theorem as well as the working.

A concise strong response distinguishes discovery from justification. Explain how the conjecture arose, but give the proof or counterexample that settles it.

A seven-session revision plan for JC2 H3 Mathematics

Session 1: Formulate a precise conjecture from three observed patterns. Session 2: Test each conjecture at boundary and negative inputs where permitted. Session 3: Find a counterexample to one false claim and explain why it is decisive.

Session 4: Prove a true pattern using two different methods. Session 5: Generalise a short arithmetic identity and discover which parameter restrictions matter. Session 6: Critique two incorrect student solutions, naming the first gap and repairing the conclusion.

Session 7: Attempt an unfamiliar mathematical passage and separate the supplied definitions, conjectures and established results. The goal is the ability to think independently rather than imitate a solved example.

A good tutor can make this learning visible by asking ‘What would falsify your claim?’ and ‘Which step converts this observation into a proof?’ rather than just saying the answer is right or wrong.

For parents considering JC2 H3 tutoring near Bukit Timah, the immutable eduKateSG Mathematics reference near Sixth Avenue MRT describes a three-learner Secondary 1 teaching model. It is not confirmation of an H3 timetable or place. Enquire directly about current JC2 H3 teaching availability.

Frequently asked questions

Does H3 Mathematics formally include mathematical investigation?

Yes. The 2026 and 2027 H3 Mathematics (9820) outlines include formulating conjectures, generalisation, special cases and completing or critiquing a solution.

How is a conjecture different from a theorem?

A conjecture is an unproved claim suggested by evidence. A theorem is established by a valid general proof.

How many counterexamples are needed to disprove a universal claim?

One valid counterexample inside the claimed domain is enough to disprove the statement.

How many numerical examples prove a universal conjecture?

No finite selection can prove a genuinely infinite universal claim without an accompanying general argument that covers all remaining cases.

What should I write when asked to critique a proof?

Identify the first invalid step, explain why it is not justified, give a counterexample if helpful, and state a correct repaired claim or proof.

Is this the same as Question 6 reading mathematical texts?

The skills overlap, but mathematical investigation also appears across unfamiliar questions. Question 6 specifically requires responding to a short mathematical text; conjecture testing and proof critique are broader habits.

Official source and useful companion guides

The official 2026 Singapore–Cambridge H3 Mathematics syllabus and 2027 syllabus identify mathematical investigation, conjecture formulation and proof critique as important expectations.

Continue with Question 6: reading mathematical texts, counterexamples and contradiction and mathematical induction. Together they help turn observations into clear, testable and defensible proofs.

JC2 H3 Mathematics Tutor Bukit Timah | Mathematical Investigation, Conjectures and Generalisation