Use the Condition to Choose the Circle
A Venn diagram makes the reference group in a conditional probability visible. In \(P(A\mid B)\), the event after the bar, \(B\), tells us to focus on the outcomes inside circle \(B\). Of those outcomes, the ones also in \(A\) lie in the overlap \(A\cap B\). So the conditional probability is the fraction of circle \(B\) that overlaps circle \(A\).
The key is not just to find the overlap. It is to compare that overlap with the whole condition group. In the earlier tutorial Conditional Probability from a Two-Way Table, the condition determined which row or column to use. A Venn diagram uses the same idea: the circle named by the condition sets the denominator.
As in The Conditional Probability Formula, when \(P(B)>0\),
The numerator is the overlap. The denominator includes both the overlap and the part of \(B\) outside \(A\), often called “B only.” It does not include the part of \(A\) outside \(B\), or the region outside both circles.
This fraction also works when the diagram shows counts rather than probabilities. Divide the count in the overlap by the total count in the condition circle. Both methods use the same reference group; only the units in the numerator and denominator differ.
One caution about diagrams: unless a problem says the areas are drawn to scale, do not estimate a probability by measuring the printed size of a circle or region. Use the probability or count labels. “Fraction of circle \(B\)” describes which outcomes make up the conditional probability, not necessarily a fraction of the diagram’s physical area.
Worked Example: Conditional Probability from Region Probabilities
Worked Example: Conditional Probability from Region Probabilities
In a fictional community survey, let \(A\) be the event that a randomly selected respondent uses a neighborhood library, and let \(B\) be the event that the respondent attends a community workshop. A Venn diagram gives \(P(A)=0.42\), \(P(B)=0.50\), and \(P(A\cap B)=0.18\). Find the probability that a respondent uses the library, given that the respondent attends a workshop.
State: We want \(P(A\mid B)\). The condition is workshop attendance, so the reference group is everyone in circle \(B\).
Plan: Use the overlap probability as the numerator and the total probability in circle \(B\) as the denominator. The condition probability is \(P(B)=0.50\), which is greater than 0, so the conditional probability is defined. To clarify the denominator, the B-only region has probability \(0.50-0.18=0.32\), and the full B circle consists of the overlap plus that region.
Do: Substitute the overlap and the probability of the condition:
As a region check, the two parts of circle \(B\) have probabilities \(0.18\) and \(0.32\), which total \(0.50\). The overlap is \(0.18\) of that \(0.50\), and \(0.18\div0.50=0.36\). The other regions are A only, \(0.42-0.18=0.24\), and neither, \(1-(0.24+0.18+0.32)=0.26\). The four regions sum to 1.
Conclude: Among community workshop attendees, the probability that a randomly selected respondent uses the neighborhood library is \(0.36\), or 36%. This is a proportion within the workshop-attendee group, not the probability of both attending a workshop and using the library in the full survey.
When the Diagram Gives Counts
With counts, the “fraction of circle \(B\)” wording can be read literally as a proportion of people or objects. Add the counts in all regions belonging to \(B\), including the overlap, to get the denominator. Then divide the overlap count by that total.
This distinction is useful because an overlap count by itself answers a different question. Dividing the overlap by the grand total gives a joint probability for a random selection from the full group. Dividing by the condition-circle total gives a conditional probability within that circle. The earlier tutorial Conditional Versus Joint Probability explains these as different reference groups.
Worked Example: A Conditional Probability from Counts
Worked Example: A Conditional Probability from Counts
Suppose a fictional group of 150 students is represented by a Venn diagram. Let \(A\) be the event that a student participates in robotics, and let \(B\) be the event that a student takes a coding course. The diagram shows 24 students in A only, 28 in the overlap, 42 in B only, and 56 in neither region. Find the probability that a student participates in robotics, given that the student takes a coding course.
State: We want \(P(A\mid B)\). The condition \(B\) means we consider only the students who take a coding course.
Plan: Count all students in circle \(B\), which includes the 28 students in both events and the 42 in B only. Use the overlap count as the numerator. The condition group contains 70 students, so it is nonempty and the conditional probability is defined.
Do: First find the size of the condition group, then divide the overlap count by it:
The full-group check is \(24+28+42+56=150\), matching the stated number of students. If we had divided the overlap by 150, we would have found \(28/150\approx0.1867\), the proportion of the full group who are in both events. That is a joint proportion, not the requested conditional probability.
Conclude: Of the students who take a coding course, 40% participate in robotics. The denominator is 70 coding-course students, not all 150 students in the group.
Changing the Condition Changes the Reference Group
The overlap stays the same if we reverse the events: \(A\cap B\) and \(B\cap A\) describe the same outcomes. But the circle used as the denominator changes. This is why \(P(A\mid B)\) and \(P(B\mid A)\) do not generally have the same value, as emphasized in Why \(P(A\mid B)\) Is Not \(P(B\mid A)\).
To read either expression, say it in words and identify the condition before calculating. “A given B” means start with B and ask what fraction is also A. “B given A” means start with A and ask what fraction is also B. The numerator is the overlap both times, but the denominators are different circle totals.
Worked Example: Reversing the Condition
Worked Example: Reversing the Condition
In a fictional survey, let \(A\) be the event that a person brings lunch from home, and let \(B\) be the event that the person uses a reusable water bottle. The Venn diagram gives \(P(A)=0.60\), \(P(B)=0.40\), and \(P(A\cap B)=0.30\). Find and compare \(P(B\mid A)\) and \(P(A\mid B)\).
First, check that the labels can describe a valid Venn diagram. The A-only probability is \(0.60-0.30=0.30\), and the B-only probability is \(0.40-0.30=0.10\). The probability of neither event is \(1-(0.30+0.30+0.10)=0.30\). These four nonnegative regions sum to 1.
Find \(P(B\mid A)\): The condition is bringing lunch, so use all of circle \(A\), with probability \(0.60\), as the denominator. The overlap has probability \(0.30\):
Find \(P(A\mid B)\): Now the condition is using a reusable bottle, so use all of circle \(B\), with probability \(0.40\), as the denominator. The overlap is still \(0.30\):
Conclude: Among people who bring lunch from home, 50% use a reusable water bottle. Among people who use a reusable water bottle, 75% bring lunch from home. The two answers differ because the condition groups have different sizes.
Common Mistakes and AP Exam Tips
- Using the grand total as the denominator. That finds a proportion of the entire group, not a conditional probability. For \(P(A\mid B)\), state that the reference group is \(B\), then use the total in circle \(B\).
- Using only the B-only region in the denominator. The condition \(B\) includes everyone in circle \(B\), including the overlap. Add the overlap and B only before dividing.
- Putting the whole condition group in the numerator. The numerator must be the outcomes satisfying both events, \(A\cap B\). The denominator is the condition group, \(B\).
- Reversing the condition without changing the denominator. The overlap remains the same, but \(P(A\mid B)\) uses \(P(B)\) below the fraction, while \(P(B\mid A)\) uses \(P(A)\).
- Reading the diagram’s drawn area as exact data. Unless the diagram is stated to be drawn to scale, use its labels and region definitions rather than estimating circle sizes.
For a clear AP response, name the condition, identify the overlap, show the fraction with the correct denominator, and interpret the result in context. A complete interpretation identifies the reference group: “Among those in \(B\), the probability of \(A\) is …” Merely saying “the probability of both events” describes the overlap, not a conditional probability.
Check Your Understanding
For each question, identify the condition circle before calculating.
- A Venn diagram has \(P(A\cap B)=0.16\) and \(P(B)=0.40\). Find \(P(A\mid B)\) and interpret it in words.
- In a group, 18 people are in the overlap of \(A\) and \(B\), and 27 are in B only. Find \(P(A\mid B)\) from the counts.
- A diagram gives \(P(A)=0.50\), \(P(B)=0.30\), and \(P(A\cap B)=0.15\). Find both \(P(A\mid B)\) and \(P(B\mid A)\). Explain why they differ or are equal.
- Explain why dividing the overlap count by the grand total does not usually give \(P(A\mid B)\).
- In \(P(A\mid B)\), which regions belong in the denominator: the overlap only, all of circle \(B\), or the entire sample space?