Population Parameter in AP Statistics

In AP Statistics, a population parameter is a fixed (usually unknown) number that describes an entire population, such as the population mean μ, proportion p, or standard deviation σ. Sample statistics estimate it, and confidence intervals and significance tests make claims about it.

What is Population Parameter?

A population parameter is a single number that describes a whole population. Think "the true proportion of all college students who work part-time" or "the mean cholesterol level of every adult in the target population." The common symbols are $\mu$ (population mean), $p$ (population proportion), and $\sigma$ or $\sigma^2$ (population standard deviation or variance).

Here's the key idea. A parameter is fixed, but you almost never know its value, because measuring an entire population is usually impossible. So you take a sample and compute a sample statistic (like $\bar{x}$ or $\hat{p}$) to estimate it. The parameter is the bullseye. The statistic is your dart. Everything in inference, from estimators to confidence intervals, is about how close your dart lands and how sure you can be about it.

Why Population Parameter matters in AP® Statistics

This term sits at the foundation of Topic 3.1 (Estimators) and Topic 3.3 (Constructing a Confidence Interval for a Population Proportion). In 3.1, the whole question is whether a statistic is a good estimator of a parameter. An estimator is unbiased when its sampling distribution is centered on the parameter's true value. In 3.3, the confidence interval exists to capture the unknown parameter $p$. If you can't name the parameter, you can't build or interpret either one. That's why AP free-response scoring so often rewards you for defining the parameter in context before doing any math. For the full procedures, head to the topic guides for 3.1 and 3.3. This page focuses on the parameter itself and how it threads through inference.

How Population Parameter connects across the course

Sample Statistic (Topic 3.1)

A statistic is the parameter's stand-in. You calculate $\hat{p}$ from your sample because you can't calculate $p$ from the population. Every pair lines up the same way: $\bar{x}$ estimates $\mu$, $\hat{p}$ estimates $p$, and $s$ estimates $\sigma$. Learn the pairs and half of inference notation stops being confusing.

Sampling Distribution and Unbiased Estimators (Topic 3.1)

The sampling distribution shows every value a statistic could take across repeated samples. The parameter is where that distribution should be centered. An estimator is unbiased when the mean of its sampling distribution equals the parameter. That's why $s^2$, computed by dividing by $n-1$, is an unbiased estimator of $\sigma^2$.

Confidence Interval (Topic 3.3)

A confidence interval is a net you throw at a fixed target. The parameter $p$ never moves. What changes from sample to sample is the interval. So when you interpret one, the claim is always about the population parameter, never about the sample proportion you already know.

Standard Error and Critical Value (Topic 3.3)

The margin of error ($\text{critical value} \times \text{standard error}$) measures how far your statistic might reasonably be from the parameter. The standard error has to use $\hat{p}$ instead of $p$, because if you knew $p$ you wouldn't need an interval at all.

Is Population Parameter on the AP® Statistics exam?

The parameter shows up on the exam in three main ways.

• Interpreting confidence intervals (MCQ and FRQ). Questions give you an interval like (0.42, 0.58) for the proportion of college students who work part-time, or (0.348, 0.492) for voters supporting Candidate A, and ask for the correct interpretation. The right answer talks about the true population proportion. Wrong choices usually make claims about the sample proportion, or say there's a 90% chance the parameter is in this specific interval.
• Unbiased estimators (MCQ). You'll be asked what makes an estimator unbiased. The answer is that the mean of its sampling distribution equals the parameter. Some questions also ask why $s^2$ is an unbiased estimator of $\sigma^2$.
• Defining the parameter in context (FRQ). Released FRQs such as 2019 Q4 (the tumbleweed/Kochia question) and 2021 Q2 (walking and cholesterol, with a random sample of 100 adults from a target population) build on the idea of a population parameter. On inference FRQs, you're expected to state the parameter clearly with its symbol, population, and variable, like "$p$ = the true proportion of all ___ who ___." A vague or sample-based definition can cost you credit.

Population Parameter vs Sample Statistic

A parameter describes the population. It's fixed and usually unknown, and it uses Greek letters or plain $p$ ($\mu$, $\sigma$, $p$). A statistic describes a sample. It's known once you collect data, it changes from sample to sample, and it usually wears a hat or bar ($\hat{p}$, $\bar{x}$, $s$). A handy memory trick is that Parameter goes with Population and Statistic goes with Sample. If a question says "42% of the 150 voters sampled," that 0.42 is a statistic, not a parameter.

Key things to remember about Population Parameter

• A population parameter is a fixed number that describes an entire population, and in practice its value is usually unknown.

• Common parameters are μ for a population mean, p for a population proportion, and σ for a population standard deviation.

• Sample statistics like x̄, p̂, and s are used to estimate parameters, and they vary from sample to sample while the parameter stays the same.

• An estimator is unbiased when the center of its sampling distribution equals the parameter it is estimating.

• A confidence interval always makes a claim about the population parameter, never about the sample statistic used to build it.

• On inference FRQs, you should define the parameter in context, naming the symbol, the population, and the variable being measured.

Frequently asked questions about Population Parameter

What is a population parameter in AP Statistics?

It's a number that describes an entire population, like the true mean μ or true proportion p. It's fixed but usually unknown, so you estimate it with a sample statistic and use confidence intervals or tests to make claims about it.

Does a population parameter change from sample to sample?

No. The parameter is fixed. What changes is the sample statistic (and any confidence interval built from it), which is exactly why statistics have sampling distributions and parameters don't.

What's the difference between a parameter and a statistic?

A parameter describes the population (μ, σ, p), and a statistic describes a sample (x̄, s, p̂). If 42% of 150 sampled voters support a candidate, 0.42 is a statistic, and the true proportion of all voters is the parameter.

Is there a 90% chance the parameter is in my confidence interval?

No, not for a specific interval you've already calculated. The parameter is either in it or not. The 90% describes the method: about 90% of intervals built this way would capture the true parameter.

What does it mean for an estimator to be unbiased for a parameter?

It means the mean of the estimator's sampling distribution equals the parameter's true value. For example, s² (dividing by n − 1) is an unbiased estimator of σ² because its sampling distribution is centered at σ².