ECON1202 Chap.5 Conditional Probability and Bayes' Formula
Conditional Probability and Bayes' Formula
Conditional probability restricts the denominator to cases where the conditioning event is already known. The multiplication law turns conditional information into a joint path probability by multiplying the probability of reaching a branch by the conditional probability that follows. A probability tree displays these roles: multiply down each path and add mutually exclusive paths that end in the same requested event.
Bayes' formula reverses conditioning. Its numerator is the joint path containing both the target cause and observed evidence, while its denominator is the total probability of that evidence across every possible cause. Independence is a separate structural claim.
Two events are independent when learning one leaves the probability of the other unchanged, equivalently when their joint probability equals the product of their marginals. Mutually exclusive positive-probability events are not independent because observing one makes the other impossible. Good solutions define events, label complements, show the evidence denominator and test probability bounds.
Weighted averages should lie between component rates, and a posterior should move in a direction consistent with how diagnostic the evidence is.
What this chapter covers
- 01
Conditional denominators
- 02
Joint probabilities
- 03
Multiplication law
- 04
Probability trees
- 05
Total probability
- 06
Bayes' reversal
- 07
Independence tests
- 08
Bounds and posterior checks
Worked example · free
Reverse a positive signal with Bayes' formula
- +1Set P(A)=0.20, P(not A)=0.80, P(S|A)=0.75 and P(S|not A)=0.10.
- +1Compute the target joint path: P(A and S)=0.20 x 0.75=0.15.
- +1Find total signal probability: P(S)=0.15+0.80 x 0.10=0.23.
- +1Divide the target path by all signal paths: P(A|S)=0.15/0.23=0.6522.
Key terms
- Conditioning event
- The event known to have occurred and used to restrict the probability denominator.
- Joint probability
- The probability that two events occur together.
- Total probability
- An event probability obtained by adding all mutually exclusive paths that lead to it.
- Bayes' formula
- A rule that reverses conditioning by weighting causes through their evidence paths.
- Independent events
- Events for which learning one does not change the probability of the other.
- Posterior probability
- The updated probability of a cause after specified evidence is observed.
Conditional Probability and Bayes' Formula FAQ
Why is P(A given B) not the same as P(B given A)?
The conditioning event determines the denominator. P(A given B) looks only inside B cases, while P(B given A) looks only inside A cases. A tree or contingency table makes the two restricted populations visibly different.
How is the Bayes denominator constructed?
List every mutually exclusive cause that could produce the observed evidence. Multiply along each cause-to-evidence path, then add those path probabilities. Omitting a path makes the posterior too large.
How can independence be tested numerically?
Compare the joint probability with the product of the marginal probabilities, or compare a conditional probability with its corresponding marginal. Equality supports independence under the stated table; inequality shows dependence.
What checks should follow a probability calculation?
Confirm every branch pair sums to one, terminal probabilities lie between zero and one, and a weighted average lies between its component rates. For Bayes, ask whether the evidence should raise or lower the prior probability.
What working should remain visible in a conditional-probability problem?
Define the events in words and symbols before substituting numbers. For a conditional probability, show the intersection in the numerator and the conditioning event in the denominator, making clear that the reference population has narrowed. On a tree, label each branch with a conditional probability, multiply along a path and add only mutually exclusive paths.
In Bayes' formula, show every route that produces the observed evidence in the denominator rather than using only the target route. For independence, test a numerical equality such as joint equals product of marginals and distinguish that concept from mutual exclusivity. Finish by checking bounds, branch sums and whether the updated probability moves in a direction consistent with the evidence.
Exam move
Define events in words before abbreviating them, then draw a two-stage tree with complements on every split. Practise the multiply-down and add-across rhythm until it is automatic. Derive Bayes from the tree rather than memorising a numerator and denominator without labels. Use contingency tables to rehearse conditional denominators and independence in a second representation.
For every problem, state which event appears after the conditioning bar and highlight the corresponding denominator population. Compare mutually exclusive and independent events using the joint-product equation so the distinction is explicit.
End with bounds and direction checks: probabilities stay between zero and one, weighted mixtures stay between their component rates, and diagnostic evidence should move the posterior toward the cause for which it is more likely. Keep event labels attached to every number to prevent a reversed answer.
Working through Conditional Probability and Bayes' Formula in ECON1202? Sia is AskSia’s AI Economics tutor — ask any ECON1202 Conditional Probability and Bayes' Formula question and get a clear, step-by-step explanation grounded in how ECON1202 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
