Sampling Distribution of the Mean

When statisticians want to learn about a population — say, the average height of all 17-year-olds in a country — they cannot measure everyone. Instead, they take a sample of, perhaps, 50 students and compute the sample mean. But if a different researcher took a different sample of 50 students, they would get a slightly different mean. The collection of all possible sample means, across every sample of the same size, forms what statisticians call the sampling distribution of the mean. The sampling distribution of the mean is not just a theoretical curiosity. It tells us exactly how much sample means vary from sample to sample, and therefore how reliable any single estimate is. Two crucial facts govern this distribution: its center equals the true population mean μ, and its spread — measured by the standard error — shrinks as sample size n increases. The larger your sample, the more trustworthy your estimate. One of the most powerful results in all of statistics, the Central Limit Theorem (CLT), states that for sufficiently large n the sampling distribution of the mean is approximately normal, regardless of the shape of the original population. This is why the normal distribution appears everywhere in statistical inference, from confidence intervals to hypothesis tests.

Key takeaways
  • Define the Population
  • Draw a Random Sample of Size n
  • Repeat Many Times
  • Identify the Mean of the Sampling Distribution
  • Calculate the Standard Error
  • Apply the Central Limit Theorem

Define the Population

Start with a population that has mean μ (the true average) and standard deviation σ (the true spread). These are fixed, unknown parameters we want to estimate. For example, suppose all exam scores in a school have μ = 72 and σ = 10.

Draw a Random Sample of Size n

Select n individuals at random. Compute their sample mean: x̄ = (x₁ + x₂ + ⋯ + xₙ) / n. This single number is one observation from the sampling distribution. It will usually be close to 72, but not exactly.

Repeat Many Times

Imagine repeating this process thousands of times — each time drawing a fresh sample of size n and recording its mean. Plot all these means as a histogram. This histogram approximates the sampling distribution of the mean. In practice, statisticians do not actually repeat sampling; instead, they use mathematical theory to describe the distribution analytically.

Identify the Mean of the Sampling Distribution

The expected value of the sample mean equals the population mean: E(x̄) = μ. This property is called unbiasedness — on average, the sample mean does not systematically overestimate or underestimate μ. Mathematically, E(x̄) = E[(x₁ + ⋯ + xₙ)/n] = (nμ)/n = μ.

Calculate the Standard Error

The standard deviation of the sampling distribution is called the standard error (SE). Because we are averaging n independent observations, the variance of x̄ is σ²/n, so SE = σ/√n. Doubling sample size reduces the standard error by a factor of √2 ≈ 1.41, not by half — a key insight. When the population SD σ is unknown, it is estimated from the sample: SE ≈ s/√n, where s is the sample standard deviation.

Apply the Central Limit Theorem

If the population is already normal, x̄ is exactly normal for any n. If not, the CLT guarantees that x̄ is approximately N(μ, σ²/n) once n is large enough — typically n ≥ 30 is used as a rule of thumb, though skewed populations may require larger samples. This allows us to compute probabilities about x̄ by converting to a Z-score: Z = (x̄ − μ) / (σ/√n).

The Key Equation

xˉ∼N ⁣(μ,  σ2n),SE=σn\bar{x} \sim N\!\left(\mu,\; \frac{\sigma^2}{n}\right), \quad \text{SE} = \frac{\sigma}{\sqrt{n}}
  • \bar{x} — Sample mean (same as data): The arithmetic average of the n sampled values; a random variable that varies across samples.
  • \mu — Population mean (same as data): The true average of all individuals in the population; the center of the sampling distribution.
  • \sigma^2 — Population variance (data units squared): The spread of individual values around μ in the full population.
  • n — Sample size (count (dimensionless)): The number of observations drawn in each sample; larger n produces a narrower sampling distribution.
  • \text{SE} — Standard error (same as data): The standard deviation of the sampling distribution of the mean; measures how much x̄ varies from sample to sample.

Worked Example

Problem. A large national test has scores with population mean μ = 500 and population standard deviation σ = 100. A school selects a random sample of n = 25 students. (a) What is the standard error of the sample mean? (b) What is the probability that the sample mean exceeds 520?

SE=20,P(xˉ>520)=P(Z>1)≈0.1587\text{SE} = 20, \quad P(\bar{x} > 520) = P(Z > 1) \approx 0.1587

Answer. Standard error = 20; P(x̄ > 520) ≈ 15.87%

Figure
Diagram
Sampling Distribution of the Mean: Effect of Sample SizeSample Mean (x̄)Densityμn = 5wide, flatn = 25n = 100narrow, tallAll curves centered at μ; SE = σ/√n shrinks as n grows
Origins · 1 minPierre-Simon Laplace and how this was discovered
PL
Pierre-Simon Laplace
1749–1827 · French

Laplace provided one of the earliest rigorous formulations of the Central Limit Theorem in 1810, showing that averages of independent random variables tend toward a normal distribution. His work in 'Théorie analytique des probabilités' (1812) laid the mathematical foundation for the entire theory of sampling distributions and remains a cornerstone of modern statistics. Laplace was so confident in the power of mathematics that he famously claimed a sufficiently powerful intellect knowing the position of every particle in the universe could predict all future events — an idea now called 'Laplace's Demon.'

  • Central Limit Theorem — The CLT is the theorem that guarantees the sampling distribution of the mean is approximately normal for large n, making all standard error calculations meaningful.
  • Standard Deviation and Variance — The standard error is derived directly from the population standard deviation σ; understanding σ is essential before studying SE = σ/√n.
  • Normal Distribution — The sampling distribution of the mean is (approximately) normal, so using it requires fluency with normal distribution properties and Z-scores.
  • Confidence Intervals — Confidence intervals for a population mean are constructed directly from the standard error of the sampling distribution, making this concept a prerequisite.