Bayes' theorem is a formula for reversing a conditional probability: P(A|B) = P(B|A) · P(A) / P(B). In AP Stats (Topic 4.5), it lets you update a probability using new information, like finding the chance someone actually has a disease given that they tested positive.
Bayes' theorem answers a sneaky question: you know P(B|A), but the problem asks for P(A|B). Those are not the same thing. The probability of a positive test given disease is not the probability of disease given a positive test, and Bayes' theorem is the machine that converts one into the other.
The formula is P(A|B) = P(B|A) · P(A) / P(B), and it comes straight from the CED's definition of conditional probability, P(A|B) = P(A ∩ B)/P(B), combined with the multiplication rule, P(A ∩ B) = P(A) · P(B|A). Just substitute the second equation into the first and you've derived Bayes' theorem yourself. On the AP exam you almost never need to memorize the formula as a formula. Instead, you build a tree diagram or a hypothetical two-way table, find the joint probability for the numerator, total up every path that leads to event B for the denominator, and divide. Same math, way less symbol-juggling.
Bayes' theorem lives in Unit 4: Probability, Random Variables, and Probability Distributions, specifically Topic 4.5: Conditional Probability. It directly supports learning objective 4.5.A (calculate conditional probabilities) and essential knowledge VAR-4.D.1 and VAR-4.D.2. The CED doesn't require you to quote the named theorem, but it absolutely requires you to do what the theorem does: combine the conditional probability definition with the multiplication rule to find a "reversed" conditional probability. This is also where probability gets real-world teeth. Diagnostic testing, quality control, and false-positive problems all hinge on the Bayes insight that a rare condition plus an imperfect test produces a surprisingly low probability that a positive result is correct.
Keep studying AP® Statistics Unit 2
Conditional probability (Unit 4)
Bayes' theorem isn't a separate idea, it's the conditional probability definition P(A|B) = P(A ∩ B)/P(B) wearing a disguise. Every Bayes problem is just a conditional probability problem where the numerator and denominator take an extra step to find. Link up to the Topic 4.5 guide for the full foundation.
Multiplication rule (Unit 4)
The multiplication rule, P(A ∩ B) = P(A) · P(B|A), supplies the numerator of Bayes' theorem. On a tree diagram, multiplying along a branch IS the multiplication rule, so every time you trace a path like P(disease) × P(positive | disease), you're building Bayes' theorem piece by piece.
Sample space (Unit 4)
Bayes problems make more sense when you picture the sample space. Conditioning on B shrinks your world to only the outcomes where B happened, and Bayes' theorem asks what fraction of that shrunken world also contains A. A tree diagram or two-way table is really just an organized map of the sample space.
Bayes-style questions show up as multiple choice and as parts of probability FRQs, almost always in one of two costumes. Costume one is the diagnostic test, where you're given sensitivity (like 90% positive given disease), specificity (like 85% negative given no disease), and a prevalence (like 2% of the population), then asked for the probability of disease given a positive test. Costume two is manufacturing, where multiple factories or processes produce items at different rates with different defect rates, and you're asked which source a defective item came from, or whether a flagged item is actually defective. Your job is the same every time. Set up a tree diagram or a hypothetical table of, say, 10,000 people, compute the joint probabilities, and divide the branch you want by the total of all branches matching the given condition. The classic trap answer is the reversed conditional, so an answer choice of 0.90 will be sitting there waiting when the real answer is something like 0.10. Show your conditional probability notation clearly, because defining events and writing P(disease | positive) correctly is part of earning credit.
P(A|B) and P(B|A) are different probabilities with different denominators, and confusing them is the single most common error in this topic. P(positive | disease) can be 0.95 while P(disease | positive) is under 0.20 when the disease is rare, because the denominator of the second one includes the huge pile of false positives from healthy people. Bayes' theorem exists precisely to convert one into the other. If your answer to a 'given a positive test' question equals the sensitivity you were handed, you computed the wrong direction.
Bayes' theorem reverses a conditional probability: it computes P(A|B) when the problem hands you P(B|A), using P(A|B) = P(B|A) · P(A) / P(B).
You can derive it yourself by plugging the multiplication rule (VAR-4.D.2) into the conditional probability definition (VAR-4.D.1), so you don't need to memorize it as a standalone formula.
On the exam, a tree diagram or a hypothetical 10,000-person table is usually faster and safer than the formula: multiply along branches, then divide the target branch by all branches matching the given condition.
When a condition is rare, even a highly accurate test gives a low probability of actually having the condition given a positive result, because false positives from the large healthy group swamp the true positives.
P(A|B) ≠ P(B|A) in general; if your final answer equals a sensitivity or specificity given in the problem, you almost certainly computed the wrong direction.
It's the formula P(A|B) = P(B|A) · P(A) / P(B), which reverses a conditional probability. In Topic 4.5, you use it to answer questions like 'given a positive test, what's the probability the person actually has the disease?'
Not as a named formula. The CED (LO 4.5.A) only requires calculating conditional probabilities, and every Bayes problem can be solved with a tree diagram or two-way table plus the basic definition P(A|B) = P(A ∩ B)/P(B).
No, and this is the biggest trap in the topic. A test with 95% sensitivity (positive given disease) can still give under a 20% chance of disease given a positive result when only 2% of people have the disease, because false positives from the healthy majority dominate the denominator.
The multiplication rule finds the probability that two events both happen: P(A ∩ B) = P(A) · P(B|A). Bayes' theorem uses that joint probability as a numerator and divides by P(B) to flip the conditioning around. The multiplication rule is an ingredient inside Bayes' theorem.
Make a tree diagram or imagine 10,000 items. With a 2% prevalence and 90% sensitivity, that's 200 diseased people with 180 true positives, plus 9,800 healthy people producing false positives at the false-positive rate. Then P(disease | positive) is just true positives divided by all positives.
Connect this key term to the AP exam workflow: review the course, practice questions, and check related study tools.
Review units, study guides, and course resources.
Check this vocabulary in multiple-choice context.
Apply key concepts in written AP responses.
Estimate the exam score you are working toward.
Review the highest-yield facts before practice.
Put the full course together before test day.