How many examples do I need before a pattern counts as proved?

No number of examples proves a conjecture. Inductive reasoning only ever gives you a statement that is probably true. Proof requires deductive reasoning from definitions, postulates, and theorems — which is exactly what two-column proofs do later in this unit. Examples are for discovering and for testing, not for proving.

How many counterexamples do I need to disprove a statement?

Exactly one, as long as it genuinely satisfies the hypothesis and fails the conclusion. A statement claims something about every case, so one failing case makes it false. Write the case explicitly with numbers or a labeled figure rather than describing it vaguely.

What is the difference between the Law of Detachment and the Law of Syllogism?

Detachment combines a conditional with a fact and produces a fact: from p arrow q and p, conclude q. Syllogism combines two conditionals and produces a new conditional: from p arrow q and q arrow r, conclude p arrow r. If your givens include a plain statement of fact, you are usually detaching; if they are all if-then statements, you are usually chaining.

Why is 'no valid conclusion' sometimes the right answer?

Because being handed the conclusion of a conditional tells you nothing about its hypothesis, and being told the hypothesis is false tells you nothing either. If the given information does not line up with either the Law of Detachment or the Law of Syllogism, the logically honest response is that no conclusion follows. Guessing a plausible-sounding statement instead is the most common error on this topic.

GEOM-2.1

Inductive & Deductive Reasoning

Learn to tell inductive from deductive reasoning, make and disprove conjectures with one counterexample, and use the Law of Detachment and Law of Syllogism.

What you'll do in this lesson

A voice-first session with the Crimsora tutor on Inductive & Deductive Reasoning, then targeted practice and FRQs — with the tutor adapting to where you get stuck.

Start Inductive & Deductive Reasoning free

What this lesson covers

Mathematicians work in two directions. Sometimes you stare at a pile of examples, notice something that keeps happening, and guess a rule — that is inductive reasoning, and the guess is called a conjecture. Other times you start from rules you already accept as true and squeeze a guaranteed conclusion out of them — that is deductive reasoning, and it is the engine of every proof you will write this year.

This lesson keeps those two moves straight. You will state conjectures from number and figure patterns, kill a false conjecture with a single well-chosen counterexample, and then apply two rules of logic — the Law of Detachment and the Law of Syllogism — to draw conclusions that are airtight. The payoff comes in the next lessons: two-column proofs are nothing but long chains of deductive steps, so the habits you build here decide how smoothly proofs go.

Inductive Reasoning and Conjectures

Inductive reasoning moves from specific observations to a general statement. You examine several cases, look for what stays the same, and state a conjecture — an unproven statement believed true based on the pattern.

Suppose you measure the angles of five different triangles and each set sums to

180^\circ

. Conjecture: the angle measures of every triangle sum to

180^\circ

. Suppose you compute

1+3=4

1+3+5=9

1+3+5+7=16

. Conjecture: the sum of the first

n

odd numbers is

n^2

.

To describe a numeric pattern, look at the differences between consecutive terms. In

3, 8, 15, 24, 35, \dots

the differences are

5, 7, 9, 11

, growing by 2 each time, so the next difference is 13 and the next term is 48. A closed formula also fits: the

n

th term is

n(n+2)

.

The crucial limitation: inductive reasoning never proves anything. It produces a good guess. Ten thousand confirming examples still leave room for case ten thousand and one to break the rule. That is why conjectures made inductively must eventually be proved deductively — or destroyed by a counterexample.

Where students go wrong: assuming the first pattern they notice is the only one. The sequence

2, 4, 8, \dots

could double each time (next term 16) or could add

2, 4, 6

(next term 14). A conjecture should be stated in words precise enough that someone else could test it, not just "the numbers go up by more each time."

Counterexamples: One Is Enough

A counterexample is a single case in which the hypothesis of a statement holds but the conclusion fails. Producing one counterexample proves the statement false, permanently. You do not need two, and you do not need to explain why the pattern broke.

Consider the conjecture: every number of the form

n^2 - n + 5

is prime. Test it. For

n=1,2,3,4

you get

5, 7, 11, 17

— all prime, which is encouraging. But

n=5

gives

25-5+5=25=5\cdot 5

, which is not prime. The conjecture is dead.

Good counterexample hunting means testing the unusual cases first, because "nice" cases are exactly the ones the pattern was built from.

Conjecture	Where to hunt	Counterexample
The square of a number is greater than the number	fractions between 0 and 1	$\left(\tfrac{1}{2}\right)^2 = \tfrac{1}{4} < \tfrac{1}{2}$
The sum of two numbers is greater than each addend	negatives and zero	$5 + (-3) = 2 < 5$
Any quadrilateral with four congruent sides is a square	non-right angles	a rhombus with $60^\circ$ angles
If $x^2 = 25$ then $x = 5$	negative roots	$x = -5$

Two frequent mistakes. First, offering a case that does not satisfy the hypothesis — a triangle is not a counterexample to a claim about quadrilaterals. Second, arguing in words that the statement "seems wrong" without a specific number or figure. A counterexample must be concrete and checkable: state the case, show the hypothesis is met, show the conclusion fails.

Deductive Reasoning and the Law of Detachment

Deductive reasoning starts with facts, definitions, postulates, or given true statements and reaches a conclusion that must be true. If the starting statements are true and the logic is valid, the conclusion is guaranteed — no testing required.

The Law of Detachment is the first tool. If the conditional

p \rightarrow q

is true and

p

is true, then

q

is true.

Given: If an angle measures

90^\circ

, then it is a right angle. Given:

m\angle B = 90^\circ

. Conclusion:

\angle B

is a right angle.

The trap is the converse error, sometimes called affirming the conclusion. Suppose you know "If a figure is a square, then it has four right angles" and you are told a figure has four right angles. You may not conclude it is a square — it could be a non-square rectangle. Detachment requires the hypothesis to be the part you are handed, not the conclusion.

Given conditional	Second given	Valid conclusion?
$p \rightarrow q$	$p$ is true	Yes: $q$ is true
$p \rightarrow q$	$q$ is true	No conclusion
$p \rightarrow q$	$p$ is false	No conclusion
$p \rightarrow q$	$q$ is false	Yes: $p$ is false

The last row is worth remembering: if the conclusion fails, the hypothesis could not have held. On homework, when the second given matches the conclusion of the conditional instead of the hypothesis, the honest answer is "no valid conclusion" — and that is often the correct response, not a trick.

The Law of Syllogism: Chaining Conditionals

The Law of Syllogism links conditionals end to end. If

p \rightarrow q

is true and

q \rightarrow r

is true, then

p \rightarrow r

is true.

Given: If a figure is a square, then it is a rhombus. Given: If a figure is a rhombus, then its diagonals are perpendicular. Conclusion: If a figure is a square, then its diagonals are perpendicular.

Notice the conclusion of the first statement is exactly the hypothesis of the second. That overlap is what makes the chain work. If the middle pieces do not match word for word in meaning, there is no chain. "If it snows, school closes" and "If the roads are icy, buses run late" share nothing, so no syllogism is possible.

Order does not matter when the statements are handed to you scrambled; matching does. Given "If

x

is divisible by 4, then

x

is even" and "If

x

is even, then

x

is not prime or

x = 2

," you rearrange so the shared statement sits in the middle before writing the chain.

Syllogism and detachment work together. Chain conditionals to build a longer conditional, then detach when you are given that the first hypothesis is true. That is exactly the structure of a geometry proof: a sequence of true conditionals (definitions and theorems) linked until the desired conclusion appears.

One more caution: a syllogism produces a conditional, not a fact. From the square example you learn what would follow if a figure is a square. You still know nothing about any particular figure until someone tells you that figure is a square.

Telling the Two Kinds of Reasoning Apart

Students mix these up because both end with a confident-sounding statement. The difference is where the confidence comes from.

Feature	Inductive	Deductive
Starting point	Specific observations, examples, measurements	Accepted facts, definitions, given conditionals
Direction	Specific to general	General to specific
Result	Conjecture (probably true)	Conclusion (must be true)
Defeated by	One counterexample	Nothing, if givens are true and logic is valid
Typical signal words	"pattern," "next term," "appears that," "in every case tested"	"given," "by definition," "therefore," "it follows that"

A quick test: ask whether the reasoning uses examples as evidence. "I measured four isosceles triangles and the base angles matched, so base angles of isosceles triangles are congruent" is inductive — it is a conjecture supported by data. "By the definition of isosceles triangle, two sides are congruent, and by the Isosceles Triangle Theorem the angles opposite them are congruent, so these base angles are congruent" is deductive.

Inductive reasoning is not bad reasoning. Nearly every theorem in geometry was first spotted as a pattern by someone sketching figures. It is how mathematics is discovered. Deductive reasoning is how mathematics is confirmed. Scientists, doctors, and detectives use both constantly: notice a pattern in the data, form a hypothesis, then reason from established principles to test it. The error to avoid is treating a conjecture as though it were proved — that is what a counterexample punishes.

Key terms

Inductive reasoning.: Reasoning that moves from specific observations or examples to a general statement; it produces a likely conclusion, not a guaranteed one.
Conjecture.: An unproven statement believed to be true based on observed patterns; it must be proved or disproved.
Counterexample.: A single specific case that satisfies the hypothesis of a statement but not its conclusion, proving the statement false.
Deductive reasoning.: Reasoning from accepted facts, definitions, postulates, or given true statements to a conclusion that must be true.
Conditional statement.: An if-then statement written $p \rightarrow q$ , where $p$ is the hypothesis and $q$ is the conclusion.
Law of Detachment.: If $p \rightarrow q$ is true and $p$ is true, then $q$ is true.
Law of Syllogism.: If $p \rightarrow q$ is true and $q \rightarrow r$ is true, then $p \rightarrow r$ is true.
Valid conclusion.: A statement that follows necessarily from the given statements by a correct rule of logic, as opposed to one that merely seems plausible.

Worked example

Part A: Write the next two terms of the sequence

3, 8, 15, 24, 35, \dots

and state a conjecture for the

n

th term. Part B: Kayla conjectures that

n^2 - n + 5

is prime for every whole number

n

. Decide whether her conjecture is true, and justify. Part C: Given (1) If a quadrilateral is a square, then it is a rhombus. (2) If a quadrilateral is a rhombus, then its diagonals are perpendicular. (3) Quadrilateral

ABCD

is a square. What can you conclude, and which laws did you use?

Part A. Find the differences:

8-3=5

15-8=7

24-15=9

35-24=11

. The differences increase by 2, so the next differences are 13 and 15. The next two terms are

35+13=48

and

48+15=63

.

For a formula, compare each term to its position:

3 = 1\cdot 3

8 = 2 \cdot 4

15 = 3 \cdot 5

24 = 4 \cdot 6

. Each term is

n(n+2)

. Check

n=5

5 \cdot 7 = 35

. Conjecture: the

n

th term is

n(n+2)

, equivalently

n^2 + 2n

. This is inductive reasoning — the formula fits every case tested, but the pattern alone does not prove it continues forever.

Part B. Test small values.

n=1

gives

1-1+5=5

, prime.

n=2

gives

4-2+5=7

, prime.

n=3

gives

9-3+5=11

, prime.

n=4

gives

16-4+5=17

, prime. So far so good, but four confirming cases prove nothing. Keep going.

n=5

gives

25-5+5=25

, and

25 = 5 \cdot 5

is not prime. The value

n=5

is a counterexample, so Kayla's conjecture is false. One counterexample is sufficient; there is no need to find others.

Part C. Statements (1) and (2) chain: the conclusion of (1), "it is a rhombus," is the hypothesis of (2). By the Law of Syllogism, if a quadrilateral is a square, then its diagonals are perpendicular.

Now use given (3).

ABCD

is a square, which is the hypothesis of the new conditional, so by the Law of Detachment the diagonals of

ABCD

are perpendicular. Both parts of Part C are deductive: the conclusion is guaranteed, not merely likely.

Practice questions

Given the true statement "If a student earns at least 90 points, then the student receives an A" and the fact "Devin received an A," which conclusion is valid?

Devin earned at least 90 points.
Devin earned fewer than 90 points.
No valid conclusion can be drawn about Devin's points.
If Devin receives an A, then Devin earned at least 90 points.

Answer: No valid conclusion can be drawn about Devin's points.

The Law of Detachment requires you to be given the hypothesis, "Devin earned at least 90 points." Instead you were given the conclusion, "Devin received an A." Reasoning backward from the conclusion assumes the converse is true, which the original statement never claims. Perhaps an A is also possible through perfect attendance or a retake. The first choice is the classic converse error, and the fourth choice simply states the converse as though it were given.

Determine whether each conjecture is true or false. If false, give a counterexample. (a) If two angles are supplementary, then one of them is obtuse. (b) The product of two negative numbers is positive.

Answer: (a) False; two right angles, each $90^\circ$ , are supplementary and neither is obtuse. (b) True.

For (a), start by looking for the boundary case rather than a typical case. Supplementary means the measures sum to

180^\circ

. Choosing

90^\circ

and

90^\circ

satisfies the hypothesis, and since obtuse means strictly greater than

90^\circ

, neither angle is obtuse — the conclusion fails, so the statement is false. That one case is a complete disproof; you do not need a second. For (b), every attempt to break it fails, since multiplying two negatives always reverses sign twice. You cannot prove it with examples, but no counterexample exists; it follows deductively from the rules of signed multiplication.

Kwame is given three true statements: (1) If a number is divisible by 6, then it is divisible by 3. (2) If a number is divisible by 12, then it is divisible by 6. (3) The number

n

is divisible by 12. Write the strongest conclusion you can and name the laws used.

Answer: $n$ is divisible by 3, using the Law of Syllogism on (2) and (1) and then the Law of Detachment with (3).

First arrange the conditionals so the middle statements match: statement (2) ends with "divisible by 6" and statement (1) begins with "divisible by 6." The Law of Syllogism gives the chained conditional: if a number is divisible by 12, then it is divisible by 3. Statement (3) supplies the hypothesis of that chain, so the Law of Detachment yields the conclusion that

n

is divisible by 3. A common slip is stopping after the syllogism and writing only the conditional; because you were also told

n

is divisible by 12, you can state a fact about

n

, which is the stronger conclusion.

FAQ

How many examples do I need before a pattern counts as proved?: No number of examples proves a conjecture. Inductive reasoning only ever gives you a statement that is probably true. Proof requires deductive reasoning from definitions, postulates, and theorems — which is exactly what two-column proofs do later in this unit. Examples are for discovering and for testing, not for proving.
How many counterexamples do I need to disprove a statement?: Exactly one, as long as it genuinely satisfies the hypothesis and fails the conclusion. A statement claims something about every case, so one failing case makes it false. Write the case explicitly with numbers or a labeled figure rather than describing it vaguely.
What is the difference between the Law of Detachment and the Law of Syllogism?: Detachment combines a conditional with a fact and produces a fact: from $p \rightarrow q$ and $p$ , conclude $q$ . Syllogism combines two conditionals and produces a new conditional: from $p \rightarrow q$ and $q \rightarrow r$ , conclude $p \rightarrow r$ . If your givens include a plain statement of fact, you are usually detaching; if they are all if-then statements, you are usually chaining.
Why is 'no valid conclusion' sometimes the right answer?: Because being handed the conclusion of a conditional tells you nothing about its hypothesis, and being told the hypothesis is false tells you nothing either. If the given information does not line up with either the Law of Detachment or the Law of Syllogism, the logically honest response is that no conclusion follows. Guessing a plausible-sounding statement instead is the most common error on this topic.

Learn this with a teacher, not a page

The Crimsora tutor teaches Inductive & Deductive Reasoning live — explaining on a whiteboard, asking you questions, and adapting to where you get stuck.

Try a live session free