AP Statistics: The Complete 2026 Guide for Texas Students & Parents (New CED, Exam Format, Scores & Study Plan)

AP® Statistics has a reputation problem in both directions. Students who dislike symbolic algebra are told it is the “easy” AP math course; students strong at algebra are told it is beneath them. The score data say something more interesting than either claim, and the course is in the middle of its biggest redesign in years — so a guide written in mid–2026 has to be precise about which exam it describes.

A note on who we are before you read further. Texas CBE™ is an independent preparation platform. We are not affiliated with, endorsed by, or connected to the College Board. AP® and Advanced Placement® are trademarks registered by the College Board, used here only to identify the exam our material prepares students for. We do not reproduce College Board exam questions, scoring rubrics, or Course and Exam Description text; our material is independently authored to the publicly published course framework. Every figure below comes from the College Board primary sources listed at the end; where College Board has not published a figure, we say so rather than estimate.

Read this first: there are two different AP Statistics exams

If you are reading this in the second half of 2026, the exam you have heard described by an older sibling, a tutor, or a three-year-old prep book is not the exam you will sit.

• The May 2026 administration — Thursday, May 7, 2026, at 12 p.m. local time — ran under the retired course description: 40 multiple-choice questions and six free-response questions, the last of which was a separately weighted Investigative Task.
• The revised Course and Exam Description is effective for the 2026–27 school year, and College Board states that the first administration of the revised exam is May 2027.

So a student enrolling in AP Statistics for fall 2026 is enrolling in the redesigned course, and the rest of this guide describes that course and exam. College Board had not published a 2027 exam calendar as of this writing, so we do not print a 2027 test date — check the official AP Statistics exam-dates page in the fall.

One practical consequence: be skeptical of study material that still teaches an Investigative Task, shows five answer choices per question, or includes a unit on inference for regression slopes. Those are artifacts of the retired version.

What the redesigned exam looks like

Per the Course and Exam Description effective Fall 2026, total testing time is three hours, split evenly between two sections:

• Section I — Multiple choice. 42 questions, 90 minutes, 50% of the exam score.
• Section II — Free response. Four questions worth 10 points each, 90 minutes, 50% of the exam score.

The four free-response questions are equally weighted, and College Board assigns each one 12.5% of the total exam score. That is a much heavier per-question stake than the retired six-question format, and it is the most important structural fact for planning preparation: losing one free-response question means losing an eighth of the exam.

The four questions are not interchangeable. Under the revised design each targets named course practices:

• Question 1 — multi-focus on Practices 1 and 2: formulating statistical questions and collecting data.
• Question 2 — multi-focus on Practices 3 and 4: analyzing data and interpreting results.
• Question 3 — inference: a hypothesis test or a confidence interval. This one is predictable, and it is the question most worth drilling to automaticity.
• Question 4 — multi-focus on Practices 2, 3, and 4, pulling collection, analysis, and interpretation into a single scenario.

Free response is scored on an analytic scale — readers award points for identifiable components, not for a general impression. That detail matters more than it sounds.

Multiple choice changed too. The count moved from 40 to 42, answer choices were cut from five to four, and two new three-question sets were added — one on probability, random variables, and probability distributions, the other on regression analysis. Four choices instead of five shifts elimination strategy: a blind guess is now a one-in-four proposition.

Fully digital from May 2027. The May 2026 exam was hybrid — questions delivered in the Bluebook application, free-response answers handwritten in paper booklets. College Board states that from May 2027 the exam is fully digital, with all student responses submitted through Bluebook and paper free-response booklets no longer used. For an exam graded substantially on written communication, that is a bigger change than it appears. Students who have practiced writing conclusions by hand should practice typing them, and should decide in advance how they will express symbols and inequalities in plain text — then stay consistent.

Calculator policy. The Course and Exam Description states that a graphing calculator with statistical capabilities is expected for both sections of the exam, and that formulas and tables are provided for both sections. There is no calculator-free portion — unlike AP Calculus, which has an explicit no-calculator section. The consequence: AP Statistics never rewards hand-computation speed. The calculator produces the number. The exam tests whether you know which number to ask for, whether the conditions permitted you to ask for it, and what it means in context.

The five units — and what those percentages actually mean

The redesigned course consolidates the material into five units. The weights published in the Course and Exam Description carry a caveat that most secondary summaries get wrong: they are labeled as approximate exam weightings for the multiple-choice section. They are not whole-exam weights.

• Unit 1 — Exploring One-Variable Data and Collecting Data: 20–30% of the multiple-choice section.
• Unit 2 — Probability, Random Variables, and Probability Distributions: 15–25%.
• Unit 3 — Inference for Categorical Data: Proportions: 15–25%.
• Unit 4 — Inference for Quantitative Data: Means: 10–20%.
• Unit 5 — Regression Analysis: 10–20%.

Read that as a study allocation and two things stand out. First, Unit 1 carries the largest single band, and it is the unit students most often treat as warm-up. Describing distributions, comparing groups, and distinguishing study designs hold the most multiple-choice real estate and, as we will see, an outsized share of avoidable free-response losses. Second, inference (Units 3 and 4 together) accounts for roughly 25–45% of the multiple-choice section and owns a guaranteed free-response question. If you are triaging, that is where the triage points.

Several topics were removed in the redesign: analyzing departures from linearity, combining random variables, the geometric distribution, the chi-square goodness-of-fit test, and the former unit on inference for regression slopes. Content on investigative statistical questions was added. And — a change with real implications for course selection — the second-year-algebra prerequisite was removed. The stated prerequisite is now first-year algebra.

The four course practices, and why they matter

Alongside the unit weights, the Course and Exam Description assigns approximate weightings to four course practices. These, too, are published as weightings for the multiple-choice section:

• Practice 1 — Formulate Questions: 5–10%.
• Practice 2 — Collect Data: 20–30%.
• Practice 3 — Analyze Data: 25–35%.
• Practice 4 — Interpret Results: 25–35%.

Be careful with these numbers, because this is where secondary summaries overreach. College Board publishes these percentages for Section I. It does not publish practice weights for Section II; what it publishes there is which practices each free-response question targets, listed earlier.

Even read conservatively, the picture is consistent. In Section I, Practice 3 — analyzing data, the part containing the arithmetic — tops out around a third. In Section II, three of the four questions explicitly target formulating, collecting, or interpreting alongside analysis. With the scoring evidence in the next section, that supports a claim which otherwise sounds like teacher folklore: AP Statistics grades communication. The deliverable is prose, and the number is an ingredient in the prose rather than the answer itself.

That is why a student who is genuinely good at math can underperform here. The habit that makes such a student fast in algebra and calculus — execute, produce the value, move on — is precisely the habit that forfeits points on this exam.

Is AP Statistics easier than AP Calculus? The honest answer

No — not by the outcome data. College Board’s published 2026 score distributions, which the score-distribution page presents as preliminary figures rounded to whole percentages, break down as follows.

• AP Statistics: 17% earned a 5, 23% a 4, 22% a 3, 17% a 2, and 21% a 1 — 62% scored 3 or higher.
• AP Calculus AB: 20% earned a 5, 28% a 4, 17% a 3, 24% a 2, and 11% a 1 — 65% scored 3 or higher.
• AP Calculus BC: 46% earned a 5, 22% a 4, 14% a 3, 14% a 2, and 4% a 1 — 82% scored 3 or higher.

Three readings a parent should take from those numbers:

• Statistics’ 3-or-higher rate sits below Calculus AB’s and far below BC’s. In 2025 the Statistics figure was about 60%, so 62% is not an anomaly.
• Statistics has the worst tail. Its 21% rate of 1s is nearly double Calculus AB’s 11% and several times BC’s 4%. It also posts the lowest 5-rate of the three.
• BC’s 82% is not evidence that BC is easy. BC draws a heavily self-selected cohort — students who have already worked through a calculus sequence and chosen to go further. Comparing raw score rates across differently selected populations tells you about who signs up at least as much as about the exam.

So the honest framing is not “harder” or “easier.” It is: Statistics has a lower barrier to entry and no lower barrier to passing. The prerequisite was just relaxed to first-year algebra, so more students can enroll. The exam then evaluates a skill — writing a defensible, contextual, precisely hedged statistical argument under time pressure — that no prerequisite screened for and most math courses never explicitly teach.

What actually costs points: evidence from the Chief Reader

College Board publishes a Chief Reader Report after each administration, summarizing how students handled the free-response questions. The 2025 AP Statistics report reads almost line for line as a catalogue of communication failures rather than mathematical ones. Note it covers Section II only; it does not report multiple-choice performance.

Summarized in our own words, the report’s guidance to teachers is to have students describe the procedure in words rather than lean on formulas, spell out “random sample” instead of abbreviating it, verify every condition and say in words that it is satisfied, report both a test statistic and a p-value, avoid calculator syntax in place of a probability, compare the p-value directly to the stated significance level, give complete context in the conclusion, and run a significance test when asked whether data provide convincing evidence.

The documented errors are worth reading closely, because they are so ordinary:

• Stating that results prove a claim, where the credited language was convincing statistical evidence to suggest it. Definitive language is a statistical error, not a stylistic one.
• Using the sample proportion in the standard error of a one-proportion test instead of the null value — and reporting a p-value from the wrong tail.
• Reversing the inequality when comparing the p-value to the significance level.
• Defining a parameter inconsistently — naming one population in the hypotheses and a different one when checking conditions.
• Failing to distinguish random sampling from random assignment — a scope-of-inference error that decides whether you may generalize, whether you may claim causation, or neither.
• Checking the large-counts condition by comparing the sample size to 30 rather than verifying that the expected successes and failures each reach 10.
• On descriptive questions: reporting both medians with no comparative language; noting an outlier without saying which distribution contained it; describing skewness without naming the direction.
• Computing an expected value correctly but never labeling the parameters — forfeiting a separate scoring component even though the arithmetic was right.

That last item is the cleanest single illustration for a skeptical parent: a correct number, unlabeled, scored lower than the same number with words around it.

How to write an inference answer that earns full credit

Because Section II is scored analytically, a full-credit inference response is essentially a checklist rendered in sentences. Train the following structure until it is automatic, then practice typing it:

• Name the procedure in words. “A one-sample z-test for a population proportion.” Not a formula, not a calculator menu item.
• State the hypotheses with the parameter defined in context. Say what the parameter represents — which population, which variable, which category counts as a success — and use standard notation.
• Check every condition, in words, with this problem’s numbers. Write “random sample” in full. Show the actual arithmetic for the large-counts check. Name the population when you invoke the 10% condition, and make sure it is the same population you named in the hypotheses.
• Report the test statistic and the p-value. Both. Written as values, not as a calculator command.
• Compare explicitly to the significance level. Write out that the p-value is less than (or greater than) α, with both numbers stated. Do not leave the comparison implied.
• Conclude with hedged language and full context. Reject or fail to reject the null hypothesis, then state that there is convincing statistical evidence about the true parameter for the named group and variable. Never “proves,” and never a bare “reject the null.”

For confidence intervals, the parallel discipline is interpreting the interval and the confidence level as separate ideas, in context, and answering the question actually asked — if the prompt asks whether there is convincing evidence, build a test, not an interval.

For descriptive questions, adopt three habits: use explicitly comparative words (“greater than,” “more variable than”), attach every feature to the distribution it belongs to, and always give shape a direction.

Statistics or Calculus — how to actually decide

Stop asking which is harder and ask which mismatch you can afford.

• Choose Calculus if the student is strong at symbolic manipulation, enjoys procedures with clean right answers, or is heading toward engineering, physics, or a mathematics-heavy major. Calculus AB still assumes precalculus; the prerequisite gap between the two courses is now wider than it was.
• Choose Statistics if the student is heading toward the life sciences, social sciences, business, public health, data analytics, or nursing — fields where an introductory statistics requirement is common — or if writing and argument are genuine strengths a purely computational course leaves unused.
• Choose both, in sequence, if the schedule allows. For an acceleration-minded student, Statistics is most often a senior-year companion to Calculus rather than a substitute. It is a different subject, not a lower rung.
• Do not choose Statistics as the easy way out. The 2026 distribution above is the counterargument, and a student who dislikes writing will find the free-response section unforgiving in a way no amount of algebra skill compensates for.

One more decision input, and it belongs to the parent: college credit policies vary by institution and by intended major. Some universities grant credit for AP Statistics toward a general requirement but not toward a quantitative major; some accept a 3, some require a 4 or 5. Check the published policy of each specific college on the student’s list before assuming any score converts to credit. No prep program, ours included, can promise a particular institution’s decision.

A realistic preparation timeline

Assume a standard fall-through-spring course with the exam in May.

August–October (Units 1–2). Build vocabulary discipline from day one. Every time you describe a distribution, say shape with direction, center, variability, and unusual features — out loud, in complete sentences. Learn the sampling-versus-assignment distinction now. Do not let “this unit is the easy one” take root; it carries the largest multiple-choice band.

November–January (Units 3–4, inference). This is the core. Write out the full inference template for every problem, even the ones you could do in your head, until the sequence is muscle memory. Keep a running error log: every point you lose, categorized as either “content” or “communication.” Most students discover the second column is longer, and that discovery is what changes behavior.

February–March (Unit 5 and consolidation). Regression, then mixed practice. Start doing full 90-minute multiple-choice sections under timing. Begin typing free-response answers rather than handwriting them, and settle your plain-text notation conventions.

April (rehearsal). At least two full three-hour rehearsals in one sitting, calculator in hand. Score your own free response against a model answer one component at a time, and count nothing as earned that a stranger could not identify in your writing.

The final week. No new content. Re-drill conditions, conclusion phrasing, and interpretation sentences. Confirm your calculator is on College Board’s approved list and has fresh batteries. Sleep.

What to do Monday, if you do nothing else: write one complete inference response — procedure named in words, hypotheses in context, conditions checked in words, statistic and p-value, explicit comparison to the significance level, hedged contextual conclusion — then grade it component by component. Repeat three times a week. That habit addresses most of what the Chief Reader flags year after year.

Dates, fees, and logistics

The 2026 AP Statistics exam was administered Thursday, May 7, 2026, at 12 p.m. local time, an afternoon slot shared with AP World History: Modern and AP African American Studies. The 2026 late-testing window ran May 18–22, 2026, with AP Statistics late testing Wednesday, May 20, 2026, at 8 a.m. local time. No 2027 calendar was published as of this writing; watch the official exam-dates page.

Fees, as published for the 2025–26 school year for exams taken in the U.S., U.S. territories, Canada, and DoDEA schools:

• Standard exam fee: $99 per exam.
• College Board fee reduction for students with significant financial need: $37 per exam, with the student responsible for the balance. Many districts and states cover part or all of the remainder — ask your counselor, as these subsidies are local.
• Late order fee: $40 per exam for exams ordered between November 15 and March 13.
• Unused or canceled exam fee: $40 per exam after the November 14 ordering deadline.
• Schools may add proctoring and administration charges.

The ordering deadline in mid-November is the one parents miss most often; put it on a calendar in September. Note the arithmetic too: at $99 per exam plus possible penalty fees, a student sitting three or four AP exams faces several hundred dollars in test fees before any preparation spending. Fees are set annually, so confirm current amounts with your AP coordinator.

Where Texas CBE fits — and what we are not

To restate plainly: Texas CBE™ is an independent preparation platform, not endorsed by, affiliated with, or connected to the College Board. AP® and Advanced Placement® are trademarks registered by the College Board. We are not the official exam and we do not administer it. We do not reproduce College Board questions, scoring rubrics, or Course and Exam Description text.

Our AP Statistics preparation is built around the failure mode this guide describes. Multiple-choice practice is aligned to the redesigned five-unit structure, and free-response practice puts the emphasis where the readers do: conditions stated in words, the explicit p-value-to-significance-level comparison, and conclusions written in the language of the problem. Free response on our platform is self-scored against a model answer — we do not grade free response. Comparing your own writing line by line against a complete model is how most students first notice they have been producing correct numbers without the sentences that earn credit.

Regular price is $39.99 per AP course for six months of access. There is a free sample quiz that requires no account — take it before deciding anything. We make no score guarantees, and no honest prep program can.

One idea to take away: in AP Statistics the number is not the answer. The sentence is the answer, and the number is evidence inside it. Students who internalize that in September do not have to relearn it in April.