Population vs sample: key differences
By Jude Wallis · Published
A population is the entire group you want to draw conclusions about; a sample is the smaller subset you actually collect data from and use to estimate the population.
AP Statistics: Unit 1 (topics 1.1 Introducing Statistics: What Can We Learn from Data?, 1.11 Random Sampling). In the Fall 2026 AP Statistics course, the population-versus-sample distinction underpins Unit 1 data collection, and the matching parameter-versus-statistic idea returns in Unit 3 topic 3.1 on estimators.
What a population is
A population is the entire collection of individuals or objects you want to draw a conclusion about. It is defined by your question, not by convenience, so if you ask about all registered voters in a state, the population is every one of them. A number that describes a population is called a parameter, written with Greek letters such as (mu, the population mean) and (sigma, the population standard deviation). You rarely get to measure a whole population, which is exactly why samples exist.
What a sample is
A sample is the subset of the population you actually collect data from. You measure the sample, compute numbers from it, and use those numbers to estimate the unknown population values. A number that describes a sample is called a statistic, written with Latin letters such as (x-bar, the sample mean) and (the sample standard deviation). A good sample is chosen at random so that it represents the population fairly; see simple random vs stratified sampling for how that is done.
Population vs sample at a glance
The two ideas differ on almost every practical axis, from the symbols you use to what you can actually observe.
| Feature | Population | Sample |
|---|---|---|
| What it is | The whole group of interest | A subset you measure |
| Size | Often huge or unreachable | Chosen to be manageable |
| Numbers from it | Parameters (fixed, usually unknown) | Statistics (vary sample to sample) |
| Mean symbol | ||
| SD symbol | ||
| Role | What you want to know | What you use to estimate it |
When you work with each
In almost every real study you work with a sample and treat the population as the target you are reasoning toward. A census, which measures every member of the population, is possible only when the group is small or you have unusual resources, so it is the exception. The whole machinery of inference exists to bridge the gap: you compute a statistic from a sample, then quantify how close it is likely to be to the parameter. See descriptive vs inferential statistics for the two jobs this involves.
The classic mix-up and how to avoid it
The most common error is confusing a parameter with a statistic, or assuming a sample result equals the population value exactly. A sample mean is only an estimate and will usually differ from the population mean by an amount called sampling error. To keep them straight, ask one question of every number: does it describe everyone in the group of interest, or only the individuals you measured? If it describes everyone it is a parameter, and if it describes just your data it is a statistic; see parameter vs statistic for more practice.
Population mean vs sample mean for a small class
A small seminar has exactly 5 students, with exam scores 80, 85, 90, 75, and 100. Treat these 5 students as the entire population. Find the population mean . Then take the sample of 3 students who scored 80, 90, and 100, and find the sample mean .
Add all 5 population scores: .
Divide by the population size 5 to get the population mean: .
Add the 3 sampled scores: .
Divide by the sample size 3 to get the sample mean: .
Compare: the sample statistic does not equal the population parameter . The 4-point gap is sampling error, the price of measuring only part of the group.
The population mean is and the sample mean is ; the statistic estimates the parameter but does not match it exactly.
Frequently asked questions
Is a sample always smaller than the population?
Yes. A sample is by definition a subset of the population, so it contains fewer individuals unless you measure everyone, in which case you have conducted a census rather than taken a sample.
Can a sample be the whole population?
When you measure every member of the population, that is a census, not a sample. A census gives you the parameters directly, so no estimation or margin of error is needed, but it is only practical for small or fully accessible groups.