Tutorials › AP Statistics › Conditional Probability and Percent Language

Conditional probability · Tutorial 271 of 1000

Conditional Probability and Percent Language

Learn to use the group named after “of” or “among” to choose the condition in a conditional probability statement.

Beginner 9 min read

What You'll Learn

  • Translate “\(x\)% of group B have characteristic A” into \(P(A\mid B)=x\%\).
  • Identify the reference group in phrases such as “of those who” and “among.”
  • Distinguish a conditional percentage from a joint percentage of the full group.
  • Explain why reversing the groups generally changes the conditional probability.
  • Convert a conditional percentage into a count or probability when group totals are given.

Percent Language Names a Reference Group

A sentence such as “40% of smokers use a particular product” gives a percentage within a specific group: smokers. It does not say that 40% of everyone uses the product. In conditional probability notation, the group after “of” supplies the condition.

This builds on the earlier tutorials What Conditional Probability Means and Reading the Given Condition in a Sentence. The condition identifies the reference group being considered. Here, the main task is to translate ordinary percent wording into notation without reversing the events.

Translation rule: “\(x\)% of people in group \(B\) have characteristic \(A\)” means that the proportion with \(A\), among the people in \(B\), is \(x\)%. In notation, \(P(A\mid B)=x/100\).

For example, let \(S\) mean that a randomly selected adult is a smoker, and let \(V\) mean that the adult uses a particular product. “40% of smokers use the product” translates to \(P(V\mid S)=0.40\). The group of smokers is the reference group; we ask what proportion of that group uses the product.

$$ \text{“40\% of smokers use the product”} \quad\longrightarrow\quad P(V\mid S)=0.40 $$

The event before the vertical bar is the characteristic whose percentage is being reported. The event after the bar is the group whose percentage is being described. Read \(P(V\mid S)\) as “the probability of product use, given that the person is a smoker.” The word “given” does not have to appear in the original sentence; “of smokers” already identifies the condition.

This order can feel counterintuitive at first. In the phrase “40% of smokers use the product,” the word “smokers” comes first, but it goes after the bar. The question being answered is “What fraction of smokers use it?” So product use is the event of interest, and smoker status defines the group.

Use “Of” and “Among” to Find the Condition

A reliable translation starts by naming the group in the denominator. Phrases such as “of the smokers,” “among smokers,” and “for people who smoke” all point to smokers as the reference group. Then identify the characteristic or outcome being reported for that group.

1
Find the reference group.
Look for the group introduced by “of,” “among,” or “given.” This will be the condition after the bar.
2
Find the reported characteristic.
Identify what the people in that group are said to do, have, or experience. This is the event before the bar.
3
Write the percentage as a proportion.
For example, 40% becomes \(0.40\), so the matching conditional probability is \(0.40\).

This method works whether the statement is about people, objects, or outcomes. “Among tablets with a cracked screen, 12% still turn on” makes cracked-screen status the condition. “Of the orders placed before noon, 85% arrived the next day” makes orders placed before noon the condition.

Formula: If \(x\)% of the members of group \(B\) have characteristic \(A\), then the reported within-group proportion is $$ P(A\mid B)=\frac{x}{100}. $$ The condition is \(B\), because the percentage is calculated within group \(B\).

When counts are available, the same translation can be written as a ratio: number in both groups divided by the number in the condition group. This is the count version of the conditional probability formula from The Conditional Probability Formula. The full-group total is not the denominator unless the sentence says the percentage is of everyone.

Worked Example: “40% of Smokers”

Worked Example: “40% of Smokers”

Suppose a fictional survey of 300 adults includes 120 smokers. Of those smokers, 48 report using a particular product. Translate “40% of smokers use the product” into conditional probability notation, and compare it with the percentage of all surveyed adults who are both smokers and product users.

State: Let \(S\) be the event that a randomly selected adult from this survey is a smoker, and let \(V\) be the event that the adult uses the product. The phrase “of smokers” identifies \(S\) as the condition.

Plan: First calculate the percentage within the smoker group by dividing the number who satisfy both events by the number of smokers. Then, separately, divide that same overlap count by the full survey total to find the proportion of all surveyed adults who satisfy both events.

Do: Among the 120 smokers, 48 use the product:

$$ P(V\mid S)=\frac{48}{120}=0.40=40\%. $$

For the full-group proportion, use all 300 surveyed adults as the denominator:

$$ P(V\cap S)=\frac{48}{300}=0.16=16\%. $$

These calculations use the same 48 adults in the numerator but different reference groups. The first denominator is 120 smokers; the second is all 300 adults. As a check, \(0.40\) of the 120 smokers is \(0.40(120)=48\), matching the stated count.

Conclude: In this fictional survey, 40% of the smokers use the product, so the within-smoker proportion is \(P(V\mid S)=0.40\). Separately, 16% of all surveyed adults are both smokers and product users. The 40% statement is conditional, not a claim that 40% of all adults use the product.

Reversing the Groups Changes the Question

The condition can be easy to reverse when a sentence names two groups. Compare “30% of tablet owners use a protective case” with “30% of protective-case users own a tablet.” The first percentage is calculated among tablet owners; the second is calculated among protective-case users. They are different questions, even though they involve the same two characteristics.

In notation, switching which group follows “of” changes the event after the bar. The overlap is the same in either question, but the reference group—and therefore the denominator—changes. This is the same distinction emphasized in the earlier tutorial Why \(P(A\mid B)\) Is Not \(P(B\mid A)\).

Worked Example: Similar Sentences, Different Conditions

Worked Example: Similar Sentences, Different Conditions

In a fictional technology survey, 80 people own a tablet, 50 people use a protective case, and 40 people do both. Let \(T\) mean that a randomly selected person owns a tablet and \(C\) mean that the person uses a protective case. Translate and calculate each statement: “50% of tablet owners use a case” and “80% of case users own a tablet.”

State: The first sentence asks for \(P(C\mid T)\), because tablet owners are the reference group. The second asks for \(P(T\mid C)\), because case users are the reference group.

Plan: For each conditional probability, use the count in both groups as the numerator. Use the size of the group named after “of” as the denominator. Both condition groups have positive counts, so each conditional probability is defined.

Do: For the first statement, there are 40 case users among 80 tablet owners:

$$ P(C\mid T)=\frac{40}{80}=0.50=50\%. $$

For the second statement, there are 40 tablet owners among 50 case users:

$$ P(T\mid C)=\frac{40}{50}=0.80=80\%. $$

The percentages also check against the counts: \(0.50(80)=40\) and \(0.80(50)=40\). The numerator is unchanged, but the two condition-group totals are 80 and 50.

Conclude: In this survey, 50% of tablet owners use a protective case, while 80% of protective-case users own a tablet. The statements have different conditions and different denominators, so the percentages need not match.

Pay Attention to “All” and “Of Those”

Sometimes a sentence contrasts a within-group percentage with a percentage of the full group. “Of the people who ordered online, 70% chose delivery” is conditional: online orders are the reference group. By contrast, “70% of all customers chose delivery” uses all customers as the reference group. The word “all” can make that full-group reference explicit.

Also distinguish “\(x\)% of group \(B\) have \(A\)” from “\(x\)% of people with \(A\) are in group \(B\).” In the first sentence, \(B\) is the condition; in the second, \(A\) is the condition. A useful habit is to complete the question: “Out of which group is this percentage being calculated?” That group belongs after the bar.

Worked Example: Read the Denominator Before Writing Notation

Worked Example: Read the Denominator Before Writing Notation

A fictional delivery service reviews 200 orders. Of 80 orders placed before noon, 68 arrived the next day. Translate “85% of orders placed before noon arrived the next day” into probability notation. Then find the percentage of all reviewed orders that were placed before noon and arrived the next day.

State: Let \(N\) mean that an order was placed before noon, and let \(D\) mean that it arrived the next day. The phrase “of orders placed before noon” makes \(N\) the condition.

Plan: The conditional percentage uses the 80 before-noon orders as its reference group. The percentage of all reviewed orders that meet both descriptions uses the grand total of 200. The count satisfying both events is 68.

Do: Within the before-noon group:

$$ P(D\mid N)=\frac{68}{80}=0.85=85\%. $$

Among all 200 reviewed orders, the proportion that were both placed before noon and delivered the next day is:

$$ P(N\cap D)=\frac{68}{200}=0.34=34\%. $$

As a count check, \(0.85(80)=68\). The conditional percentage describes delivery among the 80 early orders; the joint proportion describes the 68 qualifying orders among all 200 reviewed orders.

Conclude: The sentence “85% of orders placed before noon arrived the next day” translates to \(P(D\mid N)=0.85\). Of all reviewed orders, 34% were both placed before noon and delivered the next day.

Common Mistakes and AP Exam Tips

  • Putting the group after “of” before the bar. In “40% of smokers use the product,” smokers are the condition, so \(S\) goes after the bar: \(P(V\mid S)\), not \(P(S\mid V)\).
  • Using the full group when the statement gives a within-group percentage. “Of tablet owners” means use tablet owners as the reference group. Do not divide by everyone in the survey.
  • Confusing a conditional probability with a joint probability. “Among smokers, the proportion who use the product” is conditional. “The proportion of all adults who are smokers and use the product” is joint.
  • Assuming reversed statements are equivalent. \(P(A\mid B)\) and \(P(B\mid A)\) use the same overlap but different condition groups. Read the exact wording before assigning notation.
  • Writing a decimal but not identifying what it describes. A complete response names the group: “Among tablet owners, 50% use a case.” Saying only “the probability is 0.50” may leave the reference group unclear.

For a clear AP response, define the events, identify the group named by “of” or “among,” and put that group after the vertical bar. If the question asks for an interpretation, state the percentage in context and name the reference group. This makes clear whether the number describes a proportion within a subgroup or among everyone.

Key takeaway: In “\(x\)% of group \(B\) have characteristic \(A\),” the reported percentage is \(P(A\mid B)=x/100\). The group after “of” or “among” is the condition and sets the reference group.

Check Your Understanding

For each statement, identify the reference group and write the corresponding conditional probability.

  1. “Among people who commute by bus, 25% bring a reusable cup.” Let \(B\) mean bus commuter and \(R\) mean brings a reusable cup.
  2. “60% of repair requests submitted online were resolved within one day.” Let \(O\) mean submitted online and \(D\) mean resolved within one day.
  3. In a group of 90 people who subscribe to a service, 54 use its mobile app. What is the conditional percentage of subscribers who use the app, and how would you write it if \(S\) means subscriber and \(A\) means app user?
  4. Explain the difference between “30% of library visitors borrow a laptop” and “30% of laptop borrowers are library visitors.”
  5. A survey has 250 customers, and 40 are both online purchasers and next-day delivery users. What additional group total is needed to calculate the percentage of online purchasers who use next-day delivery?