One Overlap, Two Different Questions
In The Conditional Probability Formula, you used the probability of an overlap to calculate a conditional probability. Now compare that conditional probability with the joint probability from the very same overlap. The numerator can be the same, but the denominators differ because the two probabilities answer different questions.
A joint probability asks how likely it is that both events occur when an outcome is selected from the full group. A conditional probability asks how likely one event is among outcomes where the condition is true. The first uses the whole group as its reference group; the second restricts attention to the condition group.
The words in the question tell you which denominator to use. “Randomly selected from everyone” points to the grand total. “Among those who are in group \(B\)” points to the total in group \(B\). The overlap count appears in both calculations because both ask about outcomes that satisfy \(A\) and \(B\); what changes is the group against which that overlap is compared.
The conditional-probability count formula applies only when the count in \(B\) is greater than zero.
This count version agrees with the conditional probability formula from the previous tutorial. If all table counts come from the same group, the grand total in the probability versions cancels when you divide \(P(A\cap B)\) by \(P(B)\). The resulting ratio is the overlap count divided by the condition-group count.
Read the Reference Group Before Dividing
A useful habit is to say the requested probability in words before calculating it. For a joint probability, ask: “Out of everyone, what fraction meets both descriptions?” For a conditional probability, ask: “Within the group named after ‘given’ or ‘among,’ what fraction meets the target description?” That translation identifies the denominator.
In a two-way table, the joint probability comes from an interior cell divided by the grand total. A conditional probability also uses an interior cell, but divides by the row or column total for the condition. As in Conditional Probability from a Two-Way Table, first restrict attention to the condition group. Then count the outcomes in that group that also meet the target description.
Translate the categories in the question into events, such as \(A\) and \(B\), and identify which event is the condition if the question is conditional.
Locate the cell where both event descriptions are true. That is the numerator for the joint probability and for the conditional probability of one event given the other.
Use the grand total for a joint probability. For \(P(A\mid B)\), use the total count in \(B\), the event after the bar.
Divide, then describe the result in context. State whether it refers to everyone in the table or only to the condition group.
A quick denominator check can prevent a common error: point to the people or outcomes represented by the denominator and name that group aloud. If you cannot match the denominator to the wording of the question, pause before calculating.
Worked Example: Review Sessions and Quiz Results
Worked Example: Review Sessions and Quiz Results
The table summarizes 200 fictional students. Let \(R\) mean that a student attended a review session, and let \(P\) mean that the student passed a quiz. Find (1) the probability that a randomly selected student both attended a review session and passed, and (2) the probability that a student passed, given that the student attended a review session.
| Passed quiz | Did not pass | Total | |
|---|---|---|---|
| Attended review | 54 | 18 | 72 |
| Did not attend review | 66 | 62 | 128 |
| Total | 120 | 80 | 200 |
State: The first question asks for the joint probability \(P(R\cap P)\). The second asks for the conditional probability \(P(P\mid R)\).
Plan: Both calculations use the count 54 from the cell where attendance and passing overlap. For the joint probability, the reference group is all 200 students. For the conditional probability, the reference group is the 72 students who attended the review session.
Do:
The divisions are \(54\div200=0.27\) and \(54\div72=0.75\). As a check using probabilities, \(P(R)=72/200=0.36\), and \(0.75(0.36)=0.27\), which matches the joint probability.
Conclude: In this fictional group, the probability that a randomly selected student both attended the review session and passed the quiz is 0.27. Among students who attended the review session, the proportion who passed is 0.75.
The answers differ because they use different reference groups. The joint probability compares the 54 students in both categories with all 200 students. The conditional probability compares those same 54 students only with the 72 review attendees. Neither calculation is a substitute for the other: each answers a different question.
Worked Example: Garden Plots and Seedling Growth
Worked Example: Garden Plots and Seedling Growth
A fictional community garden records whether 150 plots received compost and whether seedlings showed strong early growth. Let \(C\) mean that a plot received compost, and let \(G\) mean that seedlings showed strong early growth. The table gives the counts. Find the probability that a randomly selected plot received compost and had strong growth, then find the probability of strong growth among plots that received compost.
| Strong growth | Not strong growth | Total | |
|---|---|---|---|
| Received compost | 42 | 18 | 60 |
| Did not receive compost | 36 | 54 | 90 |
| Total | 78 | 72 | 150 |
State: The joint probability is \(P(C\cap G)\). The conditional probability of strong growth given compost is \(P(G\mid C)\).
Plan: The overlap count is 42. Divide it by the grand total, 150, for the joint probability. Divide it by the compost-group total, 60, for the conditional probability. The condition group is nonempty, so the conditional probability is defined.
Do:
The arithmetic is \(42\div150=0.28\) and \(42\div60=0.70\). The conditional result is also the fraction \(42/60\) of compost-treated plots with strong growth. The table gives a proportion for this fictional set; these descriptive counts alone do not establish that compost caused stronger growth.
Conclude: The probability that a randomly selected plot both received compost and had strong early growth is 0.28. Among plots that received compost, 70% had strong early growth.
The same cell count, 42, appears in both numerators. Using 150 as the denominator in both parts would make the second answer another joint probability, not a probability conditional on compost. Using 60 in both parts would restrict the reference group even when the question asks about a plot selected from all 150.
Worked Example: A Sensor Alert in a Delivery Fleet
Worked Example: A Sensor Alert in a Delivery Fleet
In a fictional group of 250 delivery vehicles, 40 had a sensor alert during a route, and 30 both had an alert and required a maintenance check afterward. Let \(A\) mean that a vehicle had a sensor alert, and let \(M\) mean that it required a maintenance check. Find the joint probability of an alert and a maintenance check, and the probability of a maintenance check among vehicles with an alert.
State: The first quantity is \(P(A\cap M)\). The second is \(P(M\mid A)\).
Plan: The overlap count is 30. The joint probability uses all 250 vehicles as its reference group. The conditional probability uses only the 40 vehicles with an alert. Since \(40>0\), the condition group exists.
Do:
The calculations are \(30\div250=0.12\) and \(30\div40=0.75\). To verify their relationship, \(P(A)=40/250=0.16\), and \(0.75(0.16)=0.12\).
Conclude: In this fictional group, 12% of all vehicles both had a sensor alert and required a maintenance check. Among vehicles with a sensor alert, 75% required a maintenance check.
The phrases “of all vehicles” and “among vehicles with an alert” describe different reference groups. When interpreting a probability, include that group so the reader can tell which question the number answers.
Common Mistakes and AP Exam Tips
- Using the grand total for a conditional probability. For \(P(A\mid B)\), the denominator is the count in \(B\), not the grand total. A full-credit explanation identifies \(B\) as the condition group.
- Using a row or column total for a joint probability. A joint probability compares the overlap with everyone in the table, so its denominator is the grand total.
- Thinking the same numerator means the same probability. The overlap can be used in both calculations, but different denominators produce probabilities with different meanings.
- Leaving out the reference group in the interpretation. “The probability is 0.75” is incomplete by itself. State, for example, “Among the students who attended the review session, 75% passed the quiz.”
- Dividing the wrong cell by the right total. Make sure the numerator is the cell where both event descriptions are true. Then select the denominator based on whether the question is joint or conditional.
- Treating a conditional result as evidence of cause. A table may describe an association, but the conditional calculation by itself does not show that one event caused the other.
For a clear response, name the requested probability, show the count ratio, and interpret the result with its reference group. In a comparison, explicitly say why the denominators differ: one calculation uses the full group and the other uses the condition group.
Check Your Understanding
For each question, identify the reference group before calculating or explaining.
- A table has 240 people in total. Of these, 36 belong to both group \(A\) and group \(B\), and 90 belong to group \(B\). Find \(P(A\cap B)\) and \(P(A\mid B)\).
- In a table of 180 orders, 48 are both express orders and delivered on time. What denominator would you use for the probability that a randomly selected order is both express and on time? Explain.
- Suppose 28 of 40 members of a club attended a workshop. Describe the conditional probability of workshop attendance among club members as a count ratio. What is its denominator counting?
- A student uses the overlap count divided by the grand total to answer a question asking for \(P(A\mid B)\). Explain what that calculation actually represents and how the denominator should change.
- Why can the joint probability and conditional probability from the same table cell have different numerical values? Answer in terms of their reference groups.