Quizzes, exams, and flashcards about Statistics: AP Statistics — the four big ideas.
← AP
A calculator is permitted throughout and a formula sheet is provided, which tells you where the marks are: not in the arithmetic, but in saying what a number is allowed to mean.
Fully digital in Bluebook, 3 hours in total. Calculators are permitted, a graphing calculator with statistical capability is expected in Section II, and reference material including an equations sheet and tables is supplied.
The four free-response questions are each worth ten points and come in a fixed order. The third is always an inference question. The fourth is deliberately multi-part and spans several content areas at once — it is the one that punishes a student who learned the course as five separate units.
Forty-two multiple-choice questions in ninety minutes is a little over two minutes each, which is generous for the arithmetic and tight for the reading.
Interpretation. This is the AP subject where the calculation is most often the easy part and the sentence afterwards is where the marks go.
Three places where that bites every year. A confidence interval is a statement about a procedure across many samples, not a probability about one interval — the population mean is a fixed number, so "there is a 95 percent chance the mean is in this range" makes the wrong thing random. A p-value is a probability computed assuming the null hypothesis is true; it is not the probability that the null hypothesis is false. And a sampling distribution is the distribution of a statistic over repeated samples, which is a different object from the population and from any one sample, with a different shape and a different spread.
Alongside those sits the distinction between what random selection buys you — the right to generalise to a population — and what random assignment buys you, which is the right to say the treatment caused the difference. They are not the same permission and an experiment can have one without the other.
Two questions from each strand: resistance and skew, correlation and a regression slope read in context, sampling and experimental design, a binomial-or-geometric setting and a sampling distribution, and the two interpretation questions above. Every data set is invented and small enough to hold in your head; the classic misreading of a confidence interval appears as a distractor and the explanation says exactly why it is wrong.
The sampling-distribution question is the one worth sitting with. It gives a strongly skewed population and samples of a hundred, and its four answers are every combination of two independent mistakes — keeping the population's shape, and keeping the population's spread. Getting it right means settling shape and spread separately.
They do not train the four free-response questions, which are half the exam and where the writing happens. For those, use College Board's released questions with their scoring guidelines and real student samples.
