A conjecture is a claim waiting for a reason
Mathematics often begins before a proof exists. A learner notices a pattern, asks whether it continues, tests cases, searches for a reason and gradually turns an observation into a precise claim. That provisional claim is a conjecture. It may eventually be proved, refined or destroyed by a counterexample.
The important habit is not guessing. It is learning to move deliberately through notice → state → test → challenge → refine → prove. A good conjecture says exactly what is being claimed and under what conditions.
Start with a pattern, but do not confuse the pattern with proof
Consider the first few odd numbers: 1, 3, 5, 7. Their running sums are 1, 4, 9, 16. A learner may conjecture that the sum of the first n positive odd numbers is n². Four examples make the conjecture plausible. They do not establish it for every positive integer.
The algebra reveals the structure. The nth odd number is 2n−1. The claim can be written 1+3+5+⋯+(2n−1)=n². A geometric interpretation is also possible: each next odd number forms the new border needed to enlarge one square number into the next.
Make the claim precise
“The answer keeps becoming a square” is an observation. “For every positive integer n, the sum of the first n positive odd integers equals n²” is a mathematical conjecture. Precision identifies the objects, the operation and the domain.
Domains matter. A statement that works for positive integers may fail for real numbers, negative values or zero. Before testing a conjecture, ask: What values are allowed? What exactly is changing? What is supposed to remain true?
Test ordinary cases and dangerous cases
Do not test only friendly values. Try zero when it is allowed. Try negative values. Try the smallest permitted value. Try equal quantities, extreme ratios, boundary cases and values that make a denominator zero. The purpose of testing is partly to attack your own conjecture.
Suppose someone conjectures that x² is always greater than x. It works for x=2 and x=10. But x=1 gives equality, while x=1/2 gives 1/4<1/2. The original conjecture is false. A better question is: for which real x is x²>x? Solving x²−x>0 gives x(x−1)>0, so x<0 or x>1.
A failed conjecture can become better mathematics
When a counterexample appears, do not throw the investigation away. Ask what condition would exclude the failure without becoming arbitrary. Perhaps the claim needs positive integers instead of all real numbers. Perhaps a triangle must be right-angled. Perhaps a denominator must be non-zero. Refinement turns failure into information.
Worked investigation: consecutive integers
Take three consecutive integers: 4,5,6 have sum 15; 10,11,12 have sum 33; −2,−1,0 have sum −3. Each sum is divisible by 3. Conjecture: the sum of any three consecutive integers is divisible by 3.
Let the integers be n, n+1 and n+2. Their sum is 3n+3=3(n+1), which is a multiple of 3 for every integer n. The algebra explains why every tested case worked. The conjecture has become a proved statement.
Worked investigation: quadrilateral angles
Several drawn quadrilaterals suggest that their interior angles sum to 360°. Measuring many diagrams provides evidence, but measurement contains error and cannot cover every quadrilateral. A structural argument is stronger: draw a diagonal in a simple quadrilateral to split it into two triangles. Each triangle has angle sum 180°, giving 360° in total.
This illustrates a central move in mathematical reasoning: replace repeated measurement with a relationship that explains all permitted cases at once.
Worked investigation: even and odd structure
Examples suggest that odd + odd is even: 3+5=8, 11+17=28. Represent odd integers as 2a+1 and 2b+1. Their sum is 2a+1+2b+1=2(a+b+1), which has the form 2k for an integer k. The representation exposes the invariant structure hidden beneath the examples.
Use technology as a searchlight, not as proof
A spreadsheet, graphing tool or short program can test hundreds of cases and reveal patterns that are difficult to notice manually. This is valuable exploration. Unless the finite computation exhausts the entire permitted domain, however, it does not automatically prove a universal statement.
Technology is especially useful for finding counterexamples, plotting behaviour and comparing representations. The mathematical question remains: why should the relationship hold, and what conditions make it valid?
Conjecture laboratory
- Investigate the sum of two consecutive integers. What parity pattern do you see? State it precisely and justify it algebraically.
- Investigate n²+n for integer n. Is it always even? Explain.
- Investigate the difference between consecutive square numbers. Write a conjecture for (n+1)²−n² and prove it.
- Test the claim that the product of two numbers is always greater than either factor. Find several different types of counterexample and refine the claim.
Solutions and reasoning
Two consecutive integers n and n+1 sum to 2n+1, which is odd. For n²+n, factor to n(n+1): one of two consecutive integers is even, so their product is even. For consecutive squares, (n+1)²−n²=2n+1, so consecutive square numbers differ by consecutive odd numbers.
The product claim fails in many ways: 1/2×1/2=1/4 is smaller than either factor; 0×5=0; multiplying a positive number by 1 leaves it unchanged; negative values complicate ordinary “greater than” intuition. The failure shows why conditions must be stated rather than silently assumed.
A reusable conjecture protocol
- Generate or observe several cases.
- Describe the pattern in ordinary language.
- State a precise conjecture with its domain and conditions.
- Test ordinary, boundary and hostile cases.
- Search deliberately for a counterexample.
- If the claim survives, identify the structure that could explain it.
- Construct a proof appropriate to the level of mathematics available.
- If it fails, refine the conditions and test again.
Conjecturing turns mathematics from answer production into investigation. Continue through the Secondary Mathematics Master Index or return to the Mathematics Learning Hub.