Difficulty: Beginner | Prerequisites: None. This is the very first topic in the course.
This topic sits right at the start of any introductory statistics course. Before you can run a single test or draw a single graph, you need the vocabulary for describing what you are studying and what data you have. Population, sample, parameter, and statistic are the four words that frame every statistical investigation. If you cannot identify each one in a given scenario, every later topic (confidence intervals, hypothesis tests, regression) will feel arbitrary.
A population is the full group you care about; a sample is the subset you actually measure. A parameter is a number describing the population (usually unknown); a statistic is the corresponding number calculated from the sample. Statistics courses exist because we almost never have access to the whole population, so we use sample statistics to estimate population parameters.
Population
The entire group of individuals or observations that a study is interested in. Think of it as the "big picture" group you want to draw conclusions about.
Sample
A subset of the population that is actually observed or measured. In simple terms, this is the slice of the population you collect data from because studying every single member would be impractical or impossible.
Parameter
A numerical summary that describes some characteristic of the population. Because you rarely observe the whole population, the parameter is usually unknown. Common notation: Greek letters (e.g. μ for the population mean, σ for the population standard deviation).
Statistic
A numerical summary calculated from a sample. It serves as your best estimate of the corresponding parameter. Common notation: Roman letters (e.g. x̄ for the sample mean, s for the sample standard deviation). Think of it as the measurable stand-in for the thing you actually want to know.
A population is the group you want to learn about.
A sample is drawn from that population.
A parameter is a fixed (but usually unknown) number describing the population.
A statistic is calculated from the sample and used to estimate the parameter.
The chain runs: population → sample → statistic → inference about the parameter.
Scenario: The U.S. Department of Energy surveys approximately 900 gasoline stations weekly. On 29 December 2003, the average price of regular petrol was $1.478 per gallon.
Population: All gasoline stations in the U.S.
Sample: The approximately 900 stations actually surveyed.
Parameter: The true average price per gallon across every station in the country (unknown).
Statistic: $1.478 per gallon, the average calculated from the 900 surveyed stations.
Ask "Who or what do we ultimately care about?" That is the population.
Ask "Who or what did we actually collect data from?" That is the sample.
Ask "What number would we know if we could measure the entire population?" That is the parameter.
Ask "What number did we actually calculate from our data?" That is the statistic.
There are no formulas for this topic, but a notation convention worth memorising:
Population parameter → Greek letter (μ, σ, p)
Sample statistic → Roman letter (x̄, s, p̂)
This convention appears throughout the entire course, so locking it in now saves confusion later.
Polling companies survey a few thousand voters (sample) to predict how an entire country (population) will vote. The reported "48% approval rating" is a statistic; the actual approval across every voter is the parameter. Medical trials work the same way: test a drug on a sample of patients, then infer what would happen to the broader patient population.
Students often confuse the parameter with the statistic. The parameter belongs to the population and is usually unknown. The statistic belongs to the sample and is the number you actually compute.
A common mistake is thinking the sample must be "small." A sample can be large (e.g. 900 stations); what makes it a sample is that it is a subset, not the whole population.
Some students describe the parameter as a single data value rather than a summary measure. A parameter is a summary of the entire population (like a mean or proportion), not one individual observation.
Confusing the population with the sample frame. The population is the conceptual group of interest; the sample frame is the list from which you actually draw your sample. They are not always the same.
⚠️ Expect a scenario-based question asking you to identify the population, sample, parameter, and statistic. This is virtually guaranteed on early exams and quizzes.
⚠️ The distinction between parameter and statistic is foundational for confidence intervals and hypothesis testing later in the course. If you muddle these now, those topics will not make sense.
⚠️ Watch for tricky wording: "the average salary of 200 surveyed employees was $52,000" gives you a statistic, not a parameter, even though $52,000 sounds definitive.
True or False: A parameter is calculated from a sample. (False, it describes a population.)
True or False: A sample is always smaller than the population. (True, by definition it is a subset.)
Fill in the blank: Greek letters denote ______, while Roman letters denote ______. (Parameters; statistics.)
True or False: $1.478 per gallon, computed from 900 surveyed stations, is a parameter. (False, it is a statistic.)
Fill in the blank: The chain of reasoning runs from ______ → ______ → ______ → inference about the ______. (Population → sample → statistic → parameter.)
Q: A university surveys 500 of its 30,000 students to find out the average number of hours spent studying per week. The result is 14.2 hours. Identify the population, sample, parameter, and statistic.
A: Population: all 30,000 students at the university. Sample: the 500 students surveyed. Parameter: the true average weekly study hours across all 30,000 students (unknown). Statistic: 14.2 hours, calculated from the 500 surveyed students.
Q: A researcher wants to know the proportion of all adults in a city who exercise regularly. She interviews 1,200 adults and finds that 58% exercise regularly. What is the statistic? What is the parameter?
A: The statistic is 58% (calculated from the sample of 1,200). The parameter is the true proportion of all adults in the city who exercise regularly (unknown).
Q: Why do we use statistics to estimate parameters rather than measuring the parameter directly?
A: Measuring every member of a population is usually too expensive, too slow, or physically impossible. A well-chosen sample gives a practical, reliable estimate of the parameter at a fraction of the cost.
This connects directly to descriptive vs. inferential statistics: descriptive statistics summarise a sample (using statistics), while inferential statistics use those summaries to draw conclusions about the population (estimating parameters). It also connects to sampling methods (how you choose the sample matters enormously for whether your statistic is a good estimate of the parameter) and to confidence intervals and hypothesis tests, which formalise how much trust you can place in a sample statistic as an estimate of the population parameter.
Population vs. sample, parameter vs. statistic, population parameter, sample statistic, inferential statistics, descriptive statistics, Greek letters statistics notation, mu vs. x-bar, sigma vs. s, STAT 3021, introduction to probability and statistics, University of Minnesota, survey sampling, census vs. sample