A probability model is a description of a random process with two parts: the sample space (the list of all possible non-overlapping outcomes) and a probability assigned to each outcome, where every probability is between 0 and 1 and they all add up to 1.
A probability model is how you write down a random process in math form. It has exactly two components. First, the sample space, the set of all possible non-overlapping outcomes. Second, a probability assignment that gives each outcome a number between 0 and 1, with all the probabilities summing to 1. Roll a fair die and your model is the sample space {1, 2, 3, 4, 5, 6} with probability 1/6 attached to each face.
If all outcomes are equally likely, the model lets you compute the probability of any event E as a simple fraction: the number of outcomes in E divided by the total number of outcomes in the sample space. The model also tells you what those numbers mean. Per the CED, a probability is the relative frequency with which an event would occur in the long run if you repeated the process over and over. So P(Heads) = 0.7 doesn't mean 7 of your next 10 flips will be heads. It means that over thousands of flips, the proportion of heads settles near 0.7.
Probability models live in Topic 4.3 (Introduction to Probability) in Unit 4 and support two learning objectives. AP Stats 4.3.A asks you to calculate probabilities for events and their complements, and you can't do that without a valid model. The model is what guarantees probabilities stay between 0 and 1 and what makes the complement rule P(E') = 1 − P(E) work. AP Stats 4.3.B asks you to interpret probabilities, and the long-run relative frequency interpretation is the standard wording AP graders look for. Beyond Topic 4.3, the probability model is the skeleton for everything else in Unit 4. A probability distribution for a random variable, a binomial setting, a sampling distribution in Unit 5: all of these are just probability models with more structure layered on.
Keep studying AP Statistics Unit 4
Sample Space (Unit 4)
The sample space is one of the two parts of every probability model. If your sample space has overlapping or missing outcomes, every probability you calculate from the model is wrong before you even start.
Probability Distribution (Unit 4)
A probability distribution is a probability model where the outcomes are values of a random variable. When you build a table of X values and their probabilities in Topic 4.7, you're building a probability model with numbers as outcomes.
Complement of an Event (Unit 4)
Because a model's probabilities must sum to 1, the probability of 'not E' is automatically 1 − P(E). That's why the complement rule is often the fastest path on 'at least one' problems.
Event (Unit 4)
An event is any subset of the sample space. The model assigns the event its probability by adding up the probabilities of the outcomes inside it, which is the whole point of writing the model down.
This concept shows up mostly in multiple-choice, and the questions test whether you respect the rules of a valid model. One classic stem gives you an impossible probability, like P(A) = 1.25, and asks what must be true. The answer hinges on knowing probabilities live between 0 and 1, inclusive. Another classic gives a result like 9 heads in 10 flips of a coin with P(Heads) = 0.7 and asks for the best interpretation. The credited answer uses long-run relative frequency, not a claim about short runs. You should also be ready to (1) name the two components of a model, sample space plus probability assignment, (2) compute P(E) as favorable outcomes over total outcomes when outcomes are equally likely, and (3) recognize when the complement rule saves time. No released FRQ has asked for a probability model by name, but FRQ probability parts quietly assume you can set one up correctly.
A probability model is the general framework for any random process, with outcomes that can be anything (heads/tails, colors, categories). A probability distribution is the specific case where the outcomes are numerical values of a random variable. Every probability distribution is a probability model, but a model for drawing a card's suit isn't a distribution because suits aren't numbers. On the exam, 'distribution' signals random-variable problems (Topics 4.7+), while 'model' signals basic setup questions in Topic 4.3.
A probability model has exactly two components: a sample space of all possible non-overlapping outcomes, and a probability assigned to each outcome.
Every probability in a valid model is between 0 and 1, inclusive, and all the probabilities must sum to 1, so a value like 1.25 is impossible.
If all outcomes are equally likely, P(E) equals the number of outcomes in E divided by the total number of outcomes in the sample space.
The complement rule, P(E') = 1 − P(E), follows directly from the model's probabilities summing to 1.
Interpret a probability as a long-run relative frequency: it tells you the proportion of times the event would occur over many, many repetitions, not what happens in any short stretch.
It's a mathematical description of a random process made of two parts: a sample space listing all possible non-overlapping outcomes, and a probability for each outcome. Each probability is between 0 and 1, and together they sum to 1.
No. The CED states a probability is always between 0 and 1, inclusive. If an exam question assigns P(A) = 1.25, the model is invalid, and that's exactly what the question wants you to recognize.
A probability distribution is a probability model whose outcomes are numerical values of a random variable. A model can describe any outcomes (like card suits), but a distribution always pairs numbers with probabilities, which is what you build starting in Topic 4.7.
No. Probability is a long-run relative frequency, so 0.7 means the proportion of heads approaches 70% over a very large number of flips. Getting 9 heads in 10 flips is unusual but completely consistent with the model, since short runs vary a lot.
Use P(E') = 1 − P(E) whenever the event you don't want is easier to count, especially 'at least one' problems. It works because all probabilities in a valid model sum to 1, so whatever isn't in E must account for the rest.
Connect this key term to the AP exam workflow: review the course, practice questions, and check related study tools.
Review units, study guides, and course resources.
Check this vocabulary in multiple-choice context.
Apply key concepts in written AP responses.
Estimate the exam score you are working toward.
Review the highest-yield facts before practice.
Put the full course together before test day.