Conditional vs joint probability

By Jude Wallis · Published

A joint probability, P(A∩B)P(A \cap B), is the chance that A and B both happen, measured against every possible outcome. A conditional probability, P(A∣B)P(A \mid B), is the chance of A measured only within the outcomes where B happened. They are linked by P(A∣B)=P(A∩B)/P(B)P(A \mid B) = P(A \cap B) / P(B).

AP Statistics: Unit 2 (topics 2.2 Summary Statistics for Two Categorical Variables, 2.6 Conditional Probability). In the Fall 2026 AP Statistics course, joint and conditional relative frequencies from a two-way table are topic 2.2, Summary Statistics for Two Categorical Variables. The conditional probability formula and the general multiplication rule are topic 2.6, Conditional Probability. Both are in Unit 2, which is 15% to 25% of the multiple-choice section.

Conditional vs joint probability: the short answer

Both probabilities involve two events, and in a two-way table both use the same cell in the numerator. The difference is the denominator.

  • A joint probability divides by the whole: everyone in the table, or the full sample space. It answers "what fraction of all cases are both A and B".
  • A conditional probability divides by one group: only the cases where the given event happened. It answers "within the B cases, what fraction are A".

One formula ties them together:

P(A∣B)=P(A∩B)P(B)P(A \mid B) = \frac{P(A \cap B)}{P(B)}

So a conditional probability is a joint probability rescaled by the probability of the condition.

What a joint probability is

A joint probability is the probability that two events both occur. It is written P(A∩B)P(A \cap B) or P(A and B)P(A \text{ and } B). The symbol ∩\cap is read "and" or "intersect".

In a two-way table of counts, a joint probability is one inner cell divided by the grand total. Because every joint probability in the table shares that grand total, the joint probabilities of all the cells add to 1.

Order does not matter for a joint probability: P(A∩B)=P(B∩A)P(A \cap B) = P(B \cap A), since "A and B" and "B and A" describe the same outcomes. A joint probability can never be larger than the probability of either event alone. For how "and" differs from "or", see union vs intersection.

What a conditional probability is

A conditional probability is the probability that one event occurs given that another has occurred. It is written P(A∣B)P(A \mid B) and read "the probability of A given B". The event after the bar is the condition. It restricts attention to the outcomes where BB happened, and the probability is computed inside that smaller group.

In a two-way table, a conditional probability is one inner cell divided by its row total or its column total, whichever matches the condition. The conditional probabilities within one row, or within one column, add to 1.

Order matters here. P(A∣B)P(A \mid B) and P(B∣A)P(B \mid A) share a numerator but divide by different totals, so they are usually different numbers. That reversal is the subject of conditional probability: why order matters.

The differences side by side

FeatureJoint probabilityConditional probability
NotationP(A∩B)P(A \cap B)P(A∣B)P(A \mid B)
Question it answersHow likely are A and B togetherHow likely is A, among the B cases
Denominator in a tableGrand totalRow or column total for the condition
Order of eventsDoes not matterMatters
Signal words"and", "both""given", "among", "of those who", "if"
What adds to 1All cells of the tableThe cells in one row or one column
SizeAt most P(A)P(A) and at most P(B)P(B)At least as large as P(A∩B)P(A \cap B)

The last row follows from the formula. P(B)P(B) is at most 1, and dividing by a number that is at most 1 cannot make the result smaller, so P(A∣B)≥P(A∩B)P(A \mid B) \geq P(A \cap B) whenever P(B)>0P(B) > 0.

Moving between the two

The formula works in both directions.

To get a conditional from a joint, divide by the probability of the condition:

P(A∣B)=P(A∩B)P(B)P(A \mid B) = \frac{P(A \cap B)}{P(B)}

To get a joint from a conditional, multiply by the probability of the condition. This is the general multiplication rule:

P(A∩B)=P(B)⋅P(A∣B)P(A \cap B) = P(B) \cdot P(A \mid B)

On a tree diagram, the second-stage branches carry conditional probabilities, and multiplying along a path gives the joint probability at the end of that path. When the two events are independent, P(A∣B)P(A \mid B) equals P(A)P(A) and the joint probability becomes the simple product P(A)⋅P(B)P(A) \cdot P(B).

The classic mix-up and how to avoid it

The mix-up is reading a conditional statement as a joint one, or the reverse. "What percent of students are seniors who drive" is joint: the base is all students. "What percent of seniors drive" is conditional: the base is seniors only. The cell in the numerator is the same in both. The base changes, and the answer changes with it.

Before computing, name the group being described. If the sentence narrows the population with "among", "of the", "given", or "if", that narrowed group is the denominator and the probability is conditional. If the sentence describes everyone and asks for two traits at once, the denominator is the grand total and the probability is joint.

A quick check on the result: a conditional probability can never be smaller than the matching joint probability. If your "given" answer came out smaller than your "and" answer, the denominators are swapped. Reading the table itself is covered in how to read a two-way table.

Same cell, three different probabilities

A survey of 400 adults records housing type and pet ownership. Of 240 people who live in a house, 150 own a pet and 90 do not. Of 160 people who live in an apartment, 50 own a pet and 110 do not. For a randomly chosen respondent, find the probability of living in a house and owning a pet, the probability of owning a pet given a house, and the probability of living in a house given a pet.

  1. Find the totals. Pet owners: 150+50=200150 + 50 = 200. Grand total: 240+160=400240 + 160 = 400.

  2. Joint probability: the house-and-pet cell over the grand total, P(house∩pet)=150400=0.375P(\text{house} \cap \text{pet}) = \frac{150}{400} = 0.375.

  3. Conditional on house: the same cell over the house total, P(pet∣house)=150240=0.625P(\text{pet} \mid \text{house}) = \frac{150}{240} = 0.625.

  4. Conditional on pet: the same cell over the pet total, P(house∣pet)=150200=0.75P(\text{house} \mid \text{pet}) = \frac{150}{200} = 0.75.

  5. Check the first conditional with the formula: P(house)=240400=0.60P(\text{house}) = \frac{240}{400} = 0.60, and 0.3750.60=0.625\frac{0.375}{0.60} = 0.625.

  6. Check the second: P(pet)=200400=0.50P(\text{pet}) = \frac{200}{400} = 0.50, and 0.3750.50=0.75\frac{0.375}{0.50} = 0.75.

  7. Compare sizes. Both conditional probabilities, 0.6250.625 and 0.750.75, are larger than the joint probability 0.3750.375, as they must be.

P(house∩pet)=0.375P(\text{house} \cap \text{pet}) = 0.375, P(pet∣house)=0.625P(\text{pet} \mid \text{house}) = 0.625, and P(house∣pet)=0.75P(\text{house} \mid \text{pet}) = 0.75. All three use the cell of 150. The denominators are 400, 240, and 200.

From a conditional to a joint and back

At one company, 30% of incoming emails are spam. A link appears in 80% of spam emails and in 10% of emails that are not spam. Find the probability that a random email is spam and contains a link, then the probability that an email is spam given that it contains a link.

  1. Identify what is given: P(spam)=0.30P(\text{spam}) = 0.30, P(link∣spam)=0.80P(\text{link} \mid \text{spam}) = 0.80, and P(link∣not spam)=0.10P(\text{link} \mid \text{not spam}) = 0.10. The two 'in ... of' statements are conditional probabilities.

  2. Joint probability by the general multiplication rule: P(spam∩link)=0.30×0.80=0.24P(\text{spam} \cap \text{link}) = 0.30 \times 0.80 = 0.24.

  3. Find the other joint probability: P(not spam)=1−0.30=0.70P(\text{not spam}) = 1 - 0.30 = 0.70, so P(not spam∩link)=0.70×0.10=0.07P(\text{not spam} \cap \text{link}) = 0.70 \times 0.10 = 0.07.

  4. Add the two ways an email can contain a link: P(link)=0.24+0.07=0.31P(\text{link}) = 0.24 + 0.07 = 0.31.

  5. Reverse the condition: P(spam∣link)=P(spam∩link)P(link)=0.240.31≈0.7742P(\text{spam} \mid \text{link}) = \frac{P(\text{spam} \cap \text{link})}{P(\text{link})} = \frac{0.24}{0.31} \approx 0.7742.

  6. Compare the three numbers. P(link∣spam)=0.80P(\text{link} \mid \text{spam}) = 0.80, P(spam∩link)=0.24P(\text{spam} \cap \text{link}) = 0.24, and P(spam∣link)≈0.7742P(\text{spam} \mid \text{link}) \approx 0.7742 describe three different things.

The joint probability that an email is spam and contains a link is 0.240.24. The conditional probability that an email is spam given that it contains a link is about 0.77420.7742.

Frequently asked questions

Is a conditional probability always bigger than the joint probability?

It is never smaller. Since P(A∣B)=P(A∩B)/P(B)P(A \mid B) = P(A \cap B) / P(B) and P(B)P(B) is at most 1, the conditional is at least as large as the joint. They are equal only when P(B)=1P(B) = 1 or when the joint probability is 0.

How do I tell from the wording which one a question wants?

Find the group being described. Words such as "given", "among", "of those who", or "if" narrow the group, which signals a conditional probability. A question about two traits at once across everyone, usually with "and" or "both", wants a joint probability.

Does the order of the events matter?

For a joint probability, no: P(A∩B)P(A \cap B) and P(B∩A)P(B \cap A) are the same number. For a conditional probability, yes: P(A∣B)P(A \mid B) and P(B∣A)P(B \mid A) divide by different totals and usually differ.