New information shrinks the sample space
In 24.2 you met the phrase that changes everything: “given that…”. This lesson gives that idea its proper name — conditional probability — and shows you how to handle it from every picture the exam uses: two-way tables, Venn diagrams and the tree diagrams of 24.1. The method never changes; what changes is that you will now spot the situation instantly, whatever costume the question wears.
The idea itself is one sentence long. Before you know anything, probabilities are measured out of everyone. The moment the question tells you something has already happened — the student is a girl, the first counter was red, the person is in set B — everyone who does not fit that information vanishes, and you recalculate out of whoever is left.
The same Venn diagram, before and after “given that the person is in B”
Example 1The table shows 80 students sorted by year group and whether they play a musical instrument. Year 10: 24 play, 16 do not. Year 11: 18 play, 22 do not. A student is chosen at random. Find (a) P(plays an instrument), (b) P(plays an instrument given that the student is in Year 11).
Source 1: Venn diagrams and tables — restrict, then count
When the information arrives as counts — a filled Venn diagram or a two-way table — conditional probability is pure region-reading, exactly as you practised in 24.2. Only now you know what you are doing is called conditional probability, and you can write it down like an examiner:
From a Venn diagram
“Given B” makes the whole B circle the new total — overlap included. The numerator is the part of that circle which is also in A: the overlap A ∩ B. So with regions 9, 6, 5, 10 you get P(A | B) = 6/(6 + 5) = 6/11.
From a two-way table
“Given girl” makes the girls row the new total; “given walks” makes the walk column the new total. The numerator is one cell inside that row or column. The grand total in the corner never appears in a conditional answer.
Conditional probability explorer
Type the four region counts, then choose the “given” set and the target set — the diagram shades your new total and the fraction is built step by step. Swap given and target to watch the order matter.
Source 2: tree diagrams — the second stage was conditional all along
Here is the secret 24.1 kept from you: every time you recalculated the second-stage fractions of a without-replacement tree, you were already doing conditional probability. Take the bag from 24.1 — 5 red and 3 blue counters, two taken without replacement. The second-stage branch “red then red” carries 4/7, because once you know the first counter was red, only 4 reds remain among 7 counters. That sentence is precisely P(red 2nd | red 1st) = 4/7.
24.1’s bag again — 5 red, 3 blue, two taken without replacement
Reading the tree with your new vocabulary also explains the multiply-along rule from 24.1. Multiplying along a path is a conditional chain: the probability that both things happen is the probability of the first, times the probability of the second given the first.
Example 2A box contains 6 white and 4 black counters. Two counters are taken at random without replacement. (a) Write down P(second is white given that the first is white). (b) Find the probability that both counters are white. (c) Given that the first counter is black, find the probability that the second is also black.
Source 3: probabilities instead of counts — and why order matters
Sometimes a question gives you no population at all — just probabilities: “P(A) = 0.6, P(B) = 0.5 and P(A ∩ B) = 0.2. Find P(A | B).” The restriction idea still works; it just runs on probabilities instead of counts. The share of B-world that is also in A is the overlap divided by the whole of B:
No formula? Invent a population
Multiply every probability by 100 and pretend there are 100 people. With P(A) = 0.6, P(B) = 0.5, P(A ∩ B) = 0.2: put 20 in the overlap, then A only = 60 − 20 = 40, B only = 50 − 20 = 30, and neither = 100 − 90 = 10. Now it is a 24.2 question: P(A | B) = 20/50 = 2/5. Filling from the inside out still applies.
Example 3Events A and B are such that P(A) = 0.6, P(B) = 0.5 and P(A ∩ B) = 0.2. Find (a) P(A | B), (b) P(B | A).
P(A | B) and P(B | A) are different questions
Example 3 makes the point with numbers — 2/5 against 1/3 — but you should also feel it in words. The two conditionals share a numerator and nothing else:
The pilot test
Practically every airline pilot speaks English, so P(speaks English | pilot) is close to 1. But pick a random English speaker: the chance they fly airliners for a living, P(pilot | speaks English), is close to 0. Same overlap — the English-speaking pilots — but divided by two wildly different worlds. Whenever you finish a conditional answer, say the sentence back to yourself: “out of the given group, how many are in the target?” If the sentence sounds reversed, the fraction is reversed too.
More worked examples
Example 4In a school of 120 students, 50 study French (F), 40 study Spanish (S) and 20 study both. A student is chosen at random. Find (a) P(F | S), (b) P(S | F), and (c) explain why the answers differ although both use the same 20 students.
Example 5Paper 4 style. In a group of 60 people, 32 own a cat (C), 25 own a dog (D) and 12 own both. (a) Draw a Venn diagram. (b) Find the probability that a randomly chosen person owns neither animal. (c) Given that a person owns a dog, find the probability that they also own a cat. (d) Given that a person does not own a cat, find the probability that they own a dog.
Your turn
Practice set
- A two-way table for 40 students shows: boys — 9 walk, 11 take the bus; girls — 12 walk, 8 take the bus. Find (a) P(walks | boy) and (b) P(boy | walks).
- A Venn diagram shows 8 in A only, 4 in A ∩ B, 6 in B only and 2 in neither. Find P(A | B) and P(B | A).
- A bag holds 7 red and 3 green sweets. Two are taken at random without replacement. Write down P(second green | first green), and find P(both green).
- P(A) = 0.55, P(B) = 0.4 and P(A ∩ B) = 0.24. Find P(A | B) and P(B | A).
- Of 50 club members, 28 like tennis (T), 23 like squash (S) and 8 like neither. Find the number who like both, and hence find P(S | T).
- (a) Restrict to the boys row: 9 walkers out of 9 + 11 = 20 boys, so P(walks | boy) = 9/20. (b) Restrict to the walk column: 9 + 12 = 21 walkers, 9 of them boys, so P(boy | walks) = 9/21 = 3/7
- n(B) = 4 + 6 = 10, so P(A | B) = 4/10 = 2/5; n(A) = 8 + 4 = 12, so P(B | A) = 4/12 = 1/3
- After a green is taken, 2 greens remain among 9 sweets: P(second green | first green) = 2/9. Chain for both: 3/10 × 2/9 = 6/90 = 1/15
- P(A | B) = 0.24/0.4 = 0.6; P(B | A) = 0.24/0.55 = 24/55 = 24/55 (≈ 0.436, 3 s.f.)
- n(T ∪ S) = 50 − 8 = 42, so both = 28 + 23 − 42 = 9. Given T, the new total is 28, and 9 of those like squash: P(S | T) = 9/28
Key points to remember
- “Given that” shrinks the sample space: the words after it name your new total, and both top and bottom of the fraction must come from inside that world. Underline them.
- From a Venn diagram: P(A | B) = overlap ÷ the whole given circle. From a table: cell ÷ the given row or column total. The grand total never appears in a conditional answer.
- On a without-replacement tree, the second-stage branches are conditional probabilities — “write down” questions want the single branch, unmultiplied.
- Multiplying along a tree path is the conditional chain: P(A ∩ B) = P(A) × P(B | A).
- When only probabilities are given, restrict with P(A | B) = P(A ∩ B)/P(B) — or invent a population of 100 and read a Venn diagram. The formula is not required by the syllabus.
- Order matters: P(A | B) ≠ P(B | A) — same overlap on top, different worlds underneath.
- Do not confuse P(A ∩ B) (divide by everyone) with P(A | B) (divide by the given set).
Where next?
🎥 Take the interactive module
Probability — the full teaching module for this topic, with real past-paper questions and mark schemes.
✏️ Practise exam questions
Probability — 25 exam-style questions worth 94 marks, with full answers and the mark scheme.
Progress is saved on this device only and appears on the notes home page.