From a Chance Process to an AP-Style Answer
A discrete probability distribution is useful only when it represents the chance process correctly. In an exam question, you may need to define a random variable, construct a table of possible values and probabilities, check that the table is valid, and use it to answer a probability question. This combines skills from Probability Distribution of a Discrete Random Variable, Checking Whether a Probability Distribution Is Valid, and Notation for Random Variables.
A strong solution makes the connection between the process and the table clear. First identify what \(X\) measures and list every value it can take. Then assign probabilities, checking that the values cover all possible outcomes and that outcomes leading to the same value of \(X\) are combined. Finally, audit the distribution before using it to answer the question.
The two validity checks from Checking Whether a Probability Distribution Is Valid are essential: each probability must be between 0 and 1, inclusive, and the probabilities for all possible values must sum to 1. A table that passes only one check is not valid. If a table is invalid, do not use it as though it were a probability model; identify the problem first.
Worked Example: Build a Distribution from a Spinner
Worked Example: Build a Distribution from a Spinner
An invented game uses a spinner divided into 10 equal-sized sectors. Four sectors award 0 points, three award 2 points, two award 5 points, and one awards 8 points. Let \(X\) be the number of points awarded on one spin. Find the probability distribution of \(X\), check its validity, and find the probability of earning at least 5 points.
Define the variable and its possible values. \(X\) is the number of points awarded on one spin. The possible values are \(0,2,5,\) and \(8\); the wheel has no other point values.
Assign probabilities. Because the 10 sectors are equal in size, each is equally likely. For each value, divide the number of sectors awarding that value by 10.
| Points, \(x\) | Number of sectors | \(P(X=x)\) |
|---|---|---|
| 0 | 4 | \(4/10=0.40\) |
| 2 | 3 | \(3/10=0.30\) |
| 5 | 2 | \(2/10=0.20\) |
| 8 | 1 | \(1/10=0.10\) |
Check validity. Each entry is between 0 and 1. The total is:
The table is a valid probability distribution. Its values cover every outcome of one spin, and its probabilities account for all 10 sectors.
Find the requested probability. “At least 5 points” means \(X\geq5\), so include the values 5 and 8:
According to this model, the probability of earning at least 5 points on one spin is 0.30. Over many comparable spins, about 30% would award at least 5 points.
Notice that the probability is based on the number of sectors, not the number of distinct point values. The four possible values are not equally likely: the spinner has four sectors for 0 points but only one for 8 points. Counting categories instead of outcomes would give the wrong distribution.
Check a Table Before Using It
Exam questions may give a proposed distribution rather than asking you to build one from scratch. Treat validity as a separate task before answering questions from the table. Check the bounds on every probability, then add all probabilities. If the values describe every possible outcome but the probabilities do not sum to 1, the table cannot be used as written.
When a single probability is inconsistent and all the others are stated to be correct, you can find the value it would need to have by subtracting the known probabilities from 1. This calculation identifies a possible repair, but it does not make the original table valid. Be explicit about which version you are using.
Worked Example: Find and Use a Missing Probability
A model describes \(X\), the number of late arrivals among three scheduled shuttle pickups. A proposed table lists \(P(X=0)=0.18\), \(P(X=1)=0.32\), \(P(X=2)=0.27\), and \(P(X=3)=0.28\). Check whether the table is valid. If the first three probabilities are correct, find the corrected probability for \(X=3\), then find the probability of no more than one late arrival.
Check the proposed table. Every listed probability is between 0 and 1, but their total is:
The proposed table is invalid because the probabilities add to 1.05, not 1. Passing the individual bounds check is not enough. We should not answer the event question using the uncorrected table.
Repair the stated inconsistency. If the first three probabilities are correct and the outcomes \(0,1,2,3\) are exhaustive, the probability for \(X=3\) must make the total equal 1:
The corrected probabilities sum to \(0.18+0.32+0.27+0.23=1.00\), and each remains between 0 and 1. Thus, the corrected table is valid under the stated assumption that the first three entries are accurate.
Find the requested probability using the corrected table. “No more than one” means \(X\leq1\), which includes 0 and 1 late arrivals:
According to the corrected model, the probability of no more than one late arrival among the three pickups is 0.50. The answer depends on repairing the invalid table as specified; using the original entries would not describe a valid distribution.
Group Different Outcomes That Give the Same Value
A random variable may assign the same value to several distinct outcomes. In that case, the probability of that value is the sum of the probabilities of those outcomes. The useful exam technique is to make an outcome list or tree, write the value of \(X\) for each outcome, and group outcomes by that value. This avoids overlooking one path or counting a path twice.
If a question states that events are independent, the multiplication rule for independent events from earlier in the course can be used to find the probability of a particular combination of outcomes. The independence statement is part of the model: do not assume it merely because two checks or events are described separately.
Worked Example: Build a Distribution from Two Independent Checks
A fictional quality-control model examines two packages. Package A has probability 0.10 of being damaged, and Package B has probability 0.20 of being damaged. Assume the damage outcomes for the two packages are independent. Let \(X\) be the number of damaged packages. Build and check the distribution, then find the probability that at least one package is damaged.
State. The target is the distribution of \(X\), followed by \(P(X\geq1)\). Here \(X\) counts damaged packages among the two examined.
Plan. The possible values are \(0,1,2\). Use the stated independence to multiply the probabilities along each outcome path. Since the paths “A damaged, B not damaged” and “A not damaged, B damaged” both give \(X=1\), add their probabilities to find \(P(X=1)\).
Do. The probability that A is not damaged is \(1-0.10=0.90\), and the probability that B is not damaged is \(1-0.20=0.80\). The four outcome paths give:
| Package A | Package B | Value of \(X\) | Path probability |
|---|---|---|---|
| Not damaged | Not damaged | 0 | \((0.90)(0.80)=0.72\) |
| Damaged | Not damaged | 1 | \((0.10)(0.80)=0.08\) |
| Not damaged | Damaged | 1 | \((0.90)(0.20)=0.18\) |
| Damaged | Damaged | 2 | \((0.10)(0.20)=0.02\) |
Combine the two paths with \(X=1\): \(0.08+0.18=0.26\). Therefore, the distribution is:
| Number damaged, \(x\) | \(P(X=x)\) |
|---|---|
| 0 | 0.72 |
| 1 | 0.26 |
| 2 | 0.02 |
Check the distribution. Each probability is between 0 and 1, and:
The three values \(0,1,2\) cover all possible counts of damaged packages, so the distribution is valid.
Conclude. At least one damaged package means \(X\geq1\), so:
According to the model, the probability that at least one of the two packages is damaged is 0.28. This result uses the stated independence assumption and describes the two-package process, not a long-term rate for all packages.
Translate the Probability Question Carefully
Once a valid distribution is available, the requested event determines which rows to combine. The notation matters: \(P(X=2)\) is one exact value, while an inequality may include several values. As emphasized in Probabilities of Ranges and Inequalities, “more than” excludes its endpoint, while “at least” includes it. For integer-valued counts, writing the inequality before selecting rows is a reliable way to avoid an endpoint error.
For example, if \(X\) can be \(0,1,2,3\), then \(X>1\) includes 2 and 3, while \(X\geq1\) includes 1, 2, and 3. If the event is “at least one,” the complement approach from Using the Complement with Random Variables may also be efficient: calculate \(1-P(X=0)\). Either approach should give the same answer when applied correctly.
Common Mistakes and AP Exam Tips
- Leaving out possible values. A distribution must include every value the random variable can take, including values with small probabilities. Define \(X\) first, then check that the table covers all its possible outcomes.
- Counting categories instead of outcomes. Distinct values of \(X\) do not have to be equally likely. Use the chance process to determine each probability, as with the spinner sectors in the first example.
- Failing to combine paths. Several outcomes can produce the same value of \(X\). Add their probabilities to get \(P(X=x)\); do not list the same value as separate rows in the final distribution.
- Checking only the total. A total of 1 does not make a table valid if an entry is negative or greater than 1. Check both requirements explicitly.
- Using an invalid table anyway. If the probabilities do not sum to 1, state that the table is invalid. Only calculate from a corrected table when the question supplies enough information to justify the correction.
- Misreading an inequality. Translate the words into symbols, then identify all included values. “At least 2” is \(X\geq2\); “more than 2” is \(X>2\).
- Giving a number without context. A full-credit interpretation names the event and the chance process. For example, say “The probability that at least one of the two packages is damaged is 0.28,” rather than only “0.28.”
On an AP response, show enough work to make the construction and the probability calculation verifiable. Include the meaning of \(X\), the values and probabilities, the validity check, the inequality or exact event being calculated, and a sentence interpreting the result in context. If independence is used, identify it as a stated or justified assumption.
Key Takeaway
A distribution question is a sequence of connected tasks: represent the chance process with a random variable, account for all possible values, audit the probabilities, and use the valid table to answer the requested event. Keeping those tasks separate makes errors easier to find and explanations easier to follow.
Check Your Understanding
Use the invented distribution below for \(X\), the number of text alerts a student receives during a randomly selected study session.
| Alerts, \(x\) | \(P(X=x)\) |
|---|---|
| 0 | 0.25 |
| 1 | 0.40 |
| 2 | 0.20 |
| 3 | 0.15 |
- Check whether the table is a valid probability distribution. Show both checks.
- Find \(P(X=2)\) and interpret it in context.
- Find \(P(X\leq1)\). Which values of \(X\) are included?
- Find \(P(X\geq1)\) by adding table entries, then check your answer using the complement.
- A student says that the four values are equally likely because the table has four rows. Explain why that reasoning is incorrect.