Sample Proportion in AP Statistics

In AP Statistics, the sample proportion (p̂, read "p-hat") is the number of successes in a sample divided by the sample size, p̂ = x/n. It's a statistic that estimates the unknown population proportion p, and it's the starting point for every proportion interval and test.

What is Sample Proportion?

The sample proportion, written p̂ ("p-hat"), is the share of your sample that has the trait you care about. Count the successes (x), divide by the sample size (n), and you get p̂ = x/n. If 240 of 300 sampled households have internet access, p̂ = 240/300 = 0.80. It only makes sense for a categorical variable, where each individual either is or isn't a "success."

The big idea is that p̂ is a statistic, so it changes from sample to sample. The true population proportion p is fixed but usually unknown. Because p̂ bounces around, it has its own sampling distribution. That distribution is centered at p, with standard deviation √(p(1−p)/n), and it's approximately normal when the expected counts of successes and failures are both at least 10. Think of p̂ as your best single guess at p, and the sampling distribution as a description of how wrong that guess tends to be.

Why Sample Proportion matters in AP® Statistics

Sample proportion runs through the whole proportions strand of inference. It shows up in Topic 3.2 (Sampling Distributions for Sample Proportions), Topic 3.3 (Constructing a Confidence Interval for a Population Proportion), Topic 3.5 (Setting Up a Test for a Population Proportion), Topic 3.9 (Sampling Distributions for the Difference Between Sample Proportions), and Topic 3.10 (Constructing a Confidence Interval for the Difference Between Two Population Proportions). If you understand how p̂ behaves, the intervals and tests stop feeling like separate formulas to memorize. They're all the same move: start with p̂, measure its variability, and reason back to p. That idea of a statistic estimating a parameter, with sampling variability attached, is the core logic of AP Stats inference.

How Sample Proportion connects across the course

Sampling Distribution of p̂ (Unit 3, Topic 3.2)

Picture taking every possible sample of size n and plotting each p̂. That histogram is the sampling distribution. It's centered at p, and it gets narrower as n grows. This is why bigger samples give more precise estimates. It's also why you can use a normal model once the Large Counts condition (np ≥ 10 and n(1−p) ≥ 10) is met.

Confidence Interval for p (Unit 3, Topic 3.3)

A confidence interval is just p̂ with a margin of error wrapped around it: p̂ ± z*√(p̂(1−p̂)/n). Since you don't know p, you plug p̂ into the standard error. The interval is centered at p̂, and the confidence level sets the critical value z*.

One-Proportion z Test (Unit 3, Topic 3.5)

In a test, p̂ is the evidence you weigh against a claimed value p₀. The test statistic z = (p̂ − p₀)/√(p₀(1−p₀)/n) asks how many standard deviations your sample result sits from the claim. Notice the switch. The standard error uses p₀, not p̂, because you assume the null hypothesis is true.

Difference of Sample Proportions (Unit 3, Topics 3.9-3.10)

With two groups, the statistic becomes p̂₁ − p̂₂. Its variances add, so the standard error is √(p̂₁(1−p̂₁)/n₁ + p̂₂(1−p̂₂)/n₂). Both samples need to be independent, and each one has to pass the Large Counts check on its own.

Is Sample Proportion on the AP® Statistics exam?

You'll rarely be asked to define p̂. You'll be asked to use it. On multiple choice, a typical stem gives you raw counts and asks for an interval, a test statistic, or the center of a sampling distribution. Many practice questions compare two groups. One example has 240 of 300 households in County 1 with internet access and 280 of 400 in County 2. You compute p̂₁ = 0.80 and p̂₂ = 0.70, so the point estimate is 0.10, and then you build the interval around it. Another gives 45 of 120 patients improving on Treatment A and 60 of 130 on Treatment B, then asks for the correct 90% interval. Wrong answer choices often use counts instead of proportions, mix up which group is subtracted, or pick the wrong z*.

On free response, p̂ is the anchor of every proportion inference problem. You'll need to compute it, check conditions with it (Large Counts, the 10% condition, random sampling), plug it into the right standard error, and interpret results in context. No released FRQ has used the phrase "sample proportion" as a tested vocabulary term. Even so, proportion intervals and tests are a staple of the free-response section, and every one of them runs through p̂.

Sample Proportion vs Population Proportion (p)

The sample proportion p̂ is a statistic. You calculate it from your data, and it changes with every new sample. The population proportion p is a parameter. It's a single fixed value for the whole population, and you usually don't know it. A quick way to keep them straight is that statistics describe samples and parameters describe populations (s goes with s, p goes with p). The mix-up costs points on hypotheses, which must always be written about p, never p̂. It also matters in standard errors. Intervals use p̂, but tests use the null value p₀.

Key things to remember about Sample Proportion

• The sample proportion is p̂ = x/n, the number of successes divided by the sample size, and it only applies to categorical data.

• p̂ is a statistic that varies from sample to sample, while the population proportion p is a fixed parameter that p̂ is trying to estimate.

• The sampling distribution of p̂ is centered at p with standard deviation √(p(1−p)/n), and it is approximately normal when np and n(1−p) are both at least 10.

• Confidence intervals use p̂ in the standard error because p is unknown, but one-proportion z tests use p₀ because the null hypothesis is assumed true.

• When comparing two groups, the point estimate is p̂₁ − p̂₂, and the standard error comes from adding the two groups' variances.

• Hypotheses are always written in terms of p, never p̂, because you already know p̂ and inference is about the population.

Frequently asked questions about Sample Proportion

What is a sample proportion in AP Stats?

It's the fraction of a sample that counts as a "success," calculated as p̂ = x/n. For example, if 120 of 300 randomly selected adults exercise regularly, p̂ = 0.40. It's the statistic you use to estimate the population proportion p.

What's the difference between p-hat and p?

p̂ is the sample proportion you calculate from data, and it changes from sample to sample. p is the true population proportion, a fixed value you're trying to estimate. Hypotheses and conclusions are always about p.

Is the sample proportion always equal to the population proportion?

No. p̂ almost never equals p exactly because of sampling variability. What's true is that p̂ is an unbiased estimator, meaning its sampling distribution is centered at p. Larger samples make p̂ land closer to p on average.

Do you use p-hat or p₀ in the standard error?

It depends on the procedure. Confidence intervals use p̂ because you have no hypothesized value. One-proportion z tests use p₀ because you're calculating how unusual p̂ would be if the null hypothesis were true.

How do you find the difference in sample proportions?

Compute each group's p̂ separately, then subtract. With 40 of 200 improving in group A and 60 of 250 in group B, p̂_A = 0.20 and p̂_B = 0.24, so p̂_A − p̂_B = −0.04. Always state which group you subtracted from which so your interval or test is interpreted correctly.