Probability, trees and conditional reasoning
Introduced| English | Meaning |
|---|---|
| conditional probability/kənˈdɪʃənl ˌprɒbəˈbɪlɪti/ | The probability of an event given that another event has occurred. |
| independent/ˌɪndɪˈpendənt/ | Describing quantities, events or choices that do not depend on one another in the relationship being considered. |
| mutually exclusive/ˈmjuːtʃuːəli eksˈkluːsɪv/ | Describing events that cannot both happen at the same time. |
Ten thousand letters, September 1990
- In September 1990, Marilyn vos Savant answered a puzzle in Parade magazine. A game-show contestant picks one of three doors; one hides a car. The host, who knows where the car is, opens another door to show a goat and offers a switch. She said: always switch — you win with probability $\tfrac{2}{3}$. About 10 000 readers wrote in, many of them with PhDs, to say she was wrong. She was right.
- Conditional probability 条件概率 asks how a probability changes once we know something has happened.
In the game-show puzzle, what is the probability of winning if the contestant does not switch? Give a decimal to 3 d.p.
Difficulty 2/5 · 140 coins + a small bonus on first completion today
The first pick is right with probability $\tfrac{1}{3}$; switching wins the other $\tfrac{2}{3}$.
Tree diagrams
- Multiply along branches: $\mathrm{P}(A \cap B) = \mathrm{P}(A) \times \mathrm{P}(B \mid A)$.
- Add the ends of branches for an "or" of separate routes.
$\mathrm{P}(A) = 0.4$ and $\mathrm{P}(B \mid A) = 0.25$. Find $\mathrm{P}(A \cap B)$.
Difficulty 2/5 · 140 coins + a small bonus on first completion today
$0.4 \times 0.25$.
The conditional formula
- $\mathrm{P}(A \mid B) = \dfrac{\mathrm{P}(A \cap B)}{\mathrm{P}(B)}$: restrict to the outcomes where $B$ happened.
- $A$ and $B$ are independent 独立的 if $\mathrm{P}(A \cap B) = \mathrm{P}(A)\mathrm{P}(B)$, or equivalently $\mathrm{P}(A \mid B) = \mathrm{P}(A)$.
Probability, trees and conditional reasoning
Multiply along a tree branch and add disjoint branches
Compare the model with the worked case and explain one change.
$\mathrm{P}(A) = 0.5$, $\mathrm{P}(B) = 0.3$ and $\mathrm{P}(A \cap B) = 0.15$. $A$ and $B$ are independent.
Difficulty 1/5 · 100 coins + a small bonus on first completion today
$0.5 \times 0.3 = 0.15$.
Worked example: a positive test
1% of people have a disease. A test is positive for 95% of people with it and for 10% of people without it. Find $\mathrm{P}(\text{disease} \mid \text{positive})$.
- Branches: $\mathrm{P}(D \cap +) = 0.01 \times 0.95 = 0.0095$; $\mathrm{P}(D' \cap +) = 0.99 \times 0.10 = 0.099$.
- Total positive: $\mathrm{P}(+) = 0.0095 + 0.099 = 0.1085$.
- Condition: $\mathrm{P}(D \mid +) = \dfrac{0.0095}{0.1085} = 0.0876$ — under 9%, because healthy people are so many more.
2% have a condition. A test is positive for 90% of those with it and 5% of those without. Find $\mathrm{P}(\text{condition} \mid \text{positive})$ to 3 d.p.
Difficulty 2/5 · 140 coins + a small bonus on first completion today
$\dfrac{0.018}{0.018 + 0.049} = \dfrac{0.018}{0.067}$.
Without replacement
- A bag has 5 red and 3 blue counters. Two are taken without replacement: $\mathrm{P}(\text{both red}) = \tfrac{5}{8} \times \tfrac{4}{7} = \tfrac{5}{14}$.
- The second branch changes because the first counter is not put back.
A bag has 4 green and 6 yellow sweets. Two are eaten. Find $\mathrm{P}(\text{both green})$.
Difficulty 1/5 · 100 coins + a small bonus on first completion today
$\tfrac{4}{10} \times \tfrac{3}{9} = \tfrac{12}{90}$.
Venn diagrams
- $\mathrm{P}(A \cup B) = \mathrm{P}(A) + \mathrm{P}(B) - \mathrm{P}(A \cap B)$.
- Mutually exclusive 互斥的 events cannot both happen: $\mathrm{P}(A \cap B) = 0$.
$\mathrm{P}(A) = 0.6$, $\mathrm{P}(B) = 0.5$, $\mathrm{P}(A \cap B) = 0.3$. Find $\mathrm{P}(A \cup B)$.
Difficulty 2/5 · 140 coins + a small bonus on first completion today
$0.6 + 0.5 - 0.3$.
$\mathrm{P}(A \mid B)$ is not $\mathrm{P}(B \mid A)$. In the example, $\mathrm{P}(+ \mid D) = 0.95$ but $\mathrm{P}(D \mid +) = 0.0876$. Confusing them makes a positive test sound far more certain than it is.
Multiply along tree branches and add separate routes. $\mathrm{P}(A \mid B) = \tfrac{\mathrm{P}(A \cap B)}{\mathrm{P}(B)}$; independence means $\mathrm{P}(A \cap B) = \mathrm{P}(A)\mathrm{P}(B)$. Without replacement, later branches change — and $\mathrm{P}(A \mid B)$ and $\mathrm{P}(B \mid A)$ are different.
In a class, 60% study Physics, 50% study Chemistry and 35% study both. Find the probability that a student who studies Chemistry also studies Physics, and decide whether the two subjects are independent.
Conditional reasoning.