The AP Statistics exam tests your ability to reason statistically and communicate that reasoning clearly in writing. This guide breaks down exactly how the exam is structured, what scorers reward, and how to avoid the process mistakes that cost students points.
Use this guide alongside the free-response review resources and score calculator available on this page.
Doing well on AP Statistics is not just about knowing formulas. Scorers use point-based rubrics that reward correct method selection, proper notation, and conclusions written in context. A student who knows the right procedure but writes a vague conclusion will lose points a student who communicates clearly will earn.
40 questions in 90 minutes. Questions test all four content areas: exploring data, collecting data, probability and sampling distributions, and inference. No partial credit. Eliminate wrong answers using statistical reasoning, not intuition.
5 questions in 65 minutes, roughly 13 minutes each. Questions typically target one primary skill area each: exploring data, collecting data, probability, sampling distributions, or inference. You must show work, name procedures, check conditions, and write conclusions in context.
1 multi-part question in 25 minutes worth more than a standard Part A question. It synthesizes skills across units and requires extended statistical argumentation. Read all parts before writing so you can plan your reasoning across the whole task.
AP Statistics FRQ rubrics are built around four scoring categories: identifying the correct method, checking required conditions, executing the procedure correctly, and writing a conclusion in context. A wrong numerical answer with correct process can still earn most of the available points. A correct number with no supporting work earns very little.
Before writing anything, determine whether the question calls for a confidence interval or hypothesis test, which parameter is involved (proportion, mean, slope, difference), and how many samples or groups are present. Misidentifying the procedure is the single most costly error on inference FRQs.
Every inference procedure requires condition verification. For proportions: random, independence (10% condition), and large counts (np and n(1-p) both at least 10). For means: random, independence, and normality (CLT or stated normal population). Write the condition name and show the numerical check using values from the problem.
Write the formula or name the test, substitute values, and report the test statistic, degrees of freedom (where applicable), and p-value or interval bounds. Label every output. Do not skip steps because you used a calculator. Scorers need to see the setup, not just the answer.
Compare the p-value to alpha (typically 0.05) and state whether there is or is not convincing statistical evidence for the alternative hypothesis. Name the population and variable. Never say 'we accept H0' or 'we prove.' For intervals, interpret the interval as a range of plausible values for the population parameter.
Each FRQ sub-part maps to a rubric component. When you finish writing, re-read your response and ask: did I name the procedure, check conditions with numbers, show the calculation, and conclude in context? If any of those four moves is missing, add it before moving on.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Use this resource to practice free-response expectations, scoring moves, and evidence for Exam Skills.
Every AP Statistics FRQ response should follow a consistent process regardless of the topic. Scorers look for the same four moves on inference questions and a similar structure on data analysis and probability questions.
| Step | What scorers want to see | Common gap |
|---|---|---|
| State | Named procedure + parameter defined in context | Writing 'mu = mean' without linking to the population |
| Plan | Conditions checked with data evidence | Listing conditions without verifying them |
| Do | Correct statistic, df, and p-value shown | Skipping formula or misidentifying degrees of freedom |
| Conclude | Decision + interpretation in context | Saying 'we accept H0' or omitting context |
Each Part A question is worth 4 points. The investigative task is worth more and has a multi-part rubric. Points are awarded by component, so a student can earn partial credit by completing some steps correctly even if others are wrong.
| Score | Typical profile |
|---|---|
| 4 | All components essentially correct |
| 3 | Three E and one P, or strong holistic performance |
| 2 | Two E and two P, or mixed performance |
| 1 | One E or mostly P responses |
| 0 | No component essentially correct |
Statistical communication means every conclusion, interpretation, and description must reference the specific variables, units, and population from the problem. Generic statistical language without context is scored as partially correct at best.
| Generic (loses points) | In context (earns credit) |
|---|---|
| The mean is higher. | The mean walking distance to the nearest train station for sampled apartments is higher than expected. |
| We reject H0. | There is convincing statistical evidence that rent price and distance from a train station are linearly associated. |
| The interval is (0.41, 0.49). | We are 95% confident that the true proportion of city residents who support the tax increase is between 0.41 and 0.49. |
The investigative task is the most complex question on the exam. It is multi-part, synthesizes skills from across the course, and requires you to extend your reasoning beyond standard procedures. Students who treat it like a longer Part A question tend to miss the synthesis parts.
| Part A question | Investigative task |
|---|---|
| One primary skill area | Multiple skill areas combined |
| 13 minutes suggested | 25 minutes allocated |
| Standard procedure expected | Extension and evaluation expected |
| 4 points | Weighted more heavily in scoring |
| Term | Definition |
|---|---|
| Hypothesis Test | A statistical method used to make inferences about a population based on sample data. Involves stating H0 and Ha, checking conditions, calculating a test statistic and p-value, and writing a conclusion. On the AP exam, all four steps must be present and in context to earn full credit. |
| Confidence Interval | A range of plausible values for a population parameter calculated from sample data at a specified confidence level. On the AP exam, the conclusion must interpret the interval as a range for the true population parameter, not as a probability statement about a single sample. |
| Alpha Level (α = 0.05) | The threshold used to decide whether a p-value provides convincing evidence against H0. When p is less than α, reject H0. When p is greater than or equal to α, fail to reject H0. The alpha level must be stated explicitly in your conclusion to earn full rubric credit. |
| Least-Squares Regression Line | The line that minimizes the sum of squared residuals, written as y-hat = a + bx. On the AP exam, slope and intercept interpretations must include units, direction, and context. The slope means 'for each additional [x unit], the predicted [y variable] increases or decreases by [b units].' |
| Scatterplot | A graph displaying the relationship between two quantitative variables. On the AP exam, describing a scatterplot requires addressing direction, form, strength, and any unusual features, all in context of the actual variables shown. |
Hypothesis tests never prove or accept H0. The correct language is 'we fail to reject H0' or 'there is not convincing statistical evidence that...' Writing 'we accept H0' signals a fundamental misunderstanding of inference and will cost you the conclusion component.
Writing 'random: yes, independence: yes, large counts: yes' earns no credit for condition checking. You must show the numerical verification: state that the sample was randomly selected, confirm the sample is less than 10% of the population, and calculate np and n(1-p) with actual values.
A conclusion that says 'reject H0, p < 0.05' is incomplete. A slope interpretation that says 'for each unit increase in x, y increases by 3.2' is incomplete. Every statistical statement must reference the actual variables and population from the problem to earn full credit.
Choosing a z-test for means instead of a t-test, or running a two-sample test when the data are paired, will cost you the method component and likely the conditions component too. Before writing, confirm: one or two samples? Proportion or mean? Paired or independent?
AP Statistics scorers cannot see your calculator. If you write only the output without showing the test statistic formula, substituted values, or degrees of freedom, you may lose the procedure component even when the answer is numerically correct. Show your setup.
AP Statistics FRQ rubrics are built around four components: method identification, condition checking, procedure execution, and conclusion in context. A student who selects the wrong test but executes it correctly and concludes in context can still earn 2 or 3 out of 4 points. A student who writes only a correct numerical answer with no supporting work earns very little. The free-response review resources on this page include scored sample responses that show exactly how this plays out.
Section I MCQ questions frequently test the same reasoning skills as FRQs: identifying the correct procedure for a scenario, interpreting a confidence interval correctly, evaluating whether conditions are met, and distinguishing between observational studies and experiments. Practicing FRQ reasoning sharpens MCQ performance because both sections reward statistical thinking over formula recall.
The AP Statistics exam does not label questions by unit. A single FRQ can require you to describe a distribution, evaluate a sampling method, and interpret a confidence interval in the same prompt. The mixed-units free-response review resources on this page are specifically designed to replicate this structure, making them the closest available approximation to actual exam conditions.
Open the individual guides for Exam Skills when you want a closer review of one topic.
browse guidesPractice free-response reasoning and compare your answer with scoring guidance.
practice FRQsWatch past review streams filtered to Exam Skills when you want a video walkthrough.
open videosUse unit cheatsheets for a quick visual review after you work through the notes.
open cheatsheetsEstimate your broader AP score goal after you review the course and exam format.
open calculatorSection II has 6 free-response questions worth 50% of your score. Part A includes Questions 1-5 completed in 65 minutes. Part B is Question 6, the investigative task, completed in 25 minutes. Strong responses show work, use context from the problem, and communicate statistical reasoning clearly.
The AP Statistics exam has two sections of equal weight. Section I is 40 multiple-choice questions in 90 minutes, worth 50% of your score. Section II is 6 free-response questions in 90 minutes, also worth 50%. FRQs are scored with point-based rubrics that reward correct methods, calculations, and conclusions written in context.
FRQs draw from across all nine units. Common question types include inference procedures for proportions and means, probability and distributions, experimental design, and regression analysis. Mixed-unit questions frequently ask you to identify an appropriate method, verify conditions, carry out calculations, and interpret results in context.
The most common errors include skipping condition checks before running inference procedures, writing conclusions without context, showing only a decimal answer instead of work, and using vague language like 'the data proves' instead of precise statistical phrasing. Reviewing scored sample responses is one of the most effective ways to avoid these patterns.
For Section I, budget roughly 2 minutes per multiple-choice question. In Section II, Part A gives you about 13 minutes per question across Questions 1-5. Question 6 has its own dedicated 25-minute block. Prioritize showing complete reasoning on each FRQ rather than rushing to finish, since partial credit is available.
Fiveable has unit-specific and mixed-unit FRQ practice pages with real student responses and detailed feedback from an AP Statistics teacher. These cover units from exploring data through inference for slopes. You can find them at /ap-stats/exam-skills and linked from each unit hub.