In brief
The lecture introduces conditional probability — the probability of an event given that another event has already happened. From the definition come the multiplication rule, the law of total probability and Bayes’ theorem. The practical lesson is that even an accurate medical test can produce many false positives when a condition is rare and the prior probability is ignored.
Key points
- P(A | B) = P(A ∩ B) / P(B): the probability of A given that B has happened.
- Multiplication rule: P(A ∩ B) = P(B) · P(A | B).
- Events are independent when P(A | B) = P(A): knowing B changes nothing about A.
- Mutually exclusive events with positive probability cannot be independent.
- The law of total probability splits the sample space into hypotheses and adds their contributions.
- Bayes’ theorem updates the probability of a hypothesis after an observation.
- With 1% prevalence, a 99%-accurate test can make a positive result correct only about half the time.
- The base-rate fallacy is ignoring the prior probability of a hypothesis.
Chapters
- Why conditional probability matters
- Multiplication rule
- Independence
- Law of total probability
- Bayes’ theorem
- Medical test example
- Takeaways
Terms
- Conditional probability
- The probability of event A given that B has occurred.
- Independent events
- Events for which P(A ∩ B) = P(A) · P(B).
- Prior probability
- The probability of a hypothesis before observing new data.
- Posterior probability
- The probability of a hypothesis after taking the observation into account.
Takeaways
- Write down the prior probabilities of all hypotheses before applying Bayes’ theorem.
- Check independence from the definition rather than by intuition.
- For test problems, draw a tree for 10,000 people — the error becomes visible immediately.
Check yourself
Questions about the video
Why can a 99% test still be wrong so often?
Because the condition is rare: with 1% prevalence, 99 true positives are accompanied by roughly 99 false positives . Ignoring that base rate is the base-rate fallacy .
How do I check independence?
Use the definition: P(A ∩ B) = P(A) · P(B), or equivalently P(A | B) = P(A) .