01 / 21

The Central
Limit Theorem

Discover why the normal distribution emerges from sampling: how \(\bar{X}\) behaves as \(n\) grows, regardless of the population shape.

AP Statistics

What Is a Sampling Distribution?

Imagine you take a sample of size \(n\) from a population, compute the sample mean \(\bar{x}\), and repeat this thousands of times. The distribution of all those \(\bar{x}\) values is called the sampling distribution of the sample mean.

Population vs. Sampling Distribution

Population Distribution
The shape of individual values in the population. Can be skewed, uniform, bimodal, or any shape.
Sampling Distribution
The shape of \(\bar{x}\) values across many samples. Becomes approximately normal as \(n\) increases.

Why Does the CLT Matter?

In the real world, we rarely know the population distribution. The CLT gives us permission to use normal-based methods (z-scores, confidence intervals, hypothesis tests) even when the population is not normal.

1
Confidence intervals for \(\mu\) rely on the normality of \(\bar{X}\).
2
Hypothesis tests use z or t statistics that assume \(\bar{X}\) is approximately normal.
3
Quality control and polling both depend on the CLT for valid inference.

The Central Limit Theorem: Formal Statement

Mean of \(\bar{X}\)
\(\mu_{\bar{X}} = \mu\)
Std Dev of \(\bar{X}\)
\(\sigma_{\bar{X}} = \dfrac{\sigma}{\sqrt{n}}\)
Shape of \(\bar{X}\)
Approximately Normal

From Skewed to Normal

Population (skewed) \(n = 1\) Sampling dist (\(n=10\)) \(n = 10\) Sampling dist (\(n=30\)) \(n = 30\) As \(n\) increases, sampling distribution becomes more Normal

A filling line that deposits one statistic for a whole tray

A filling line loads bottles into trays while one statistic is recorded for each tray.A tray contains individual observations, but its average contributes one value to a distribution of sample means. Repeating the sampling process reveals how that statistic varies.

Build the distribution of tray means

Twenty-six possible sample means appear as discrete probability bars with a normal approximation drawn for comparison.For 25 independent fills equally likely to be 40 or 60 mL, K is binomial with n = 25 and p = 0.5, and the tray mean is 40 + 0.8K. The exact distribution is discrete, centered at 50 mL with standard deviation 2 mL. A normal curve approximates its shape.

The Standard Error of \(\bar{X}\)

The standard deviation of the sampling distribution has a special name: the standard error. It measures how much \(\bar{x}\) typically varies from \(\mu\).

\(\mu\) \(n = 5\) \(n = 30\) larger SE smaller SE

CLT vs. Law of Large Numbers

Central Limit Theorem Law of Large Numbers
What it says \(\bar{X}\) is approximately Normal for large \(n\) \(\bar{X} \to \mu\) as \(n \to \infty\)
Focus Shape of the distribution Value of the average
Tells us about Spread and symmetry of \(\bar{X}\) Convergence of \(\bar{X}\) to \(\mu\)
Key use Constructing z-scores, CIs Justifying long-run frequencies

Conditions for Applying the CLT

1
Random: The sample must be selected randomly (SRS or proper randomization).
2
Independent: Observations must be independent. For sampling without replacement, the population must be at least 10 times the sample size (10% condition).
3
Large enough \(n\): The rule of thumb is \(n \geq 30\). If the population is nearly Normal, smaller \(n\) may suffice. If heavily skewed, use larger \(n\).

How to Apply the CLT

1
Identify the population mean \(\mu\) and standard deviation \(\sigma\).
2
Compute the standard error: \(\text{SE} = \dfrac{\sigma}{\sqrt{n}}\).
3
Check conditions: Random? Independent (10% rule)? Large sample (\(n \geq 30\))?
4
Model: \(\bar{X} \sim N\!\left(\mu,\;\dfrac{\sigma}{\sqrt{n}}\right)\).
5
Find probabilities using the z-score: \(z = \dfrac{\bar{x} - \mu}{\sigma / \sqrt{n}}\).

Example 1: Light Bulb Lifetimes

A factory produces light bulbs with a mean lifetime of \(\mu = 1200\) hours and standard deviation \(\sigma = 200\) hours. The distribution of lifetimes is right-skewed. If we take a random sample of \(n = 50\) bulbs, what is \(P(\bar{X} > 1240)\)?

Example 1: Visualized

\(\mu = 1200\) \(\bar{x} = 1240\) 0.0793 \(\text{SE} = 28.28\) \(z = 1.41\)

Example 2: Commute Times

The mean commute time in a city is \(\mu = 35\) minutes with \(\sigma = 12\) minutes. A researcher samples \(n = 64\) commuters. Find the probability that the sample mean is between 33 and 37 minutes.

Example 3: Finding Sample Size

A quality inspector wants \(P(\bar{X}\) is within 5 units of \(\mu) \geq 0.95\). The population has \(\sigma = 20\). How large must \(n\) be?

Common Mistakes

✕ Using \(\sigma\) instead of SE
When finding probabilities for \(\bar{X}\), divide by \(\sigma/\!\sqrt{n}\), not \(\sigma\). Forgetting the \(\sqrt{n}\) is the most common error.
✕ Applying CLT with small \(n\)
For highly skewed populations, \(n = 10\) is not enough. The rule of thumb is \(n \geq 30\); check the shape first.
✕ CLT makes data Normal
The CLT applies to the sampling distribution of \(\bar{X}\), not to individual observations. The population stays its original shape.

Summary: CLT at a Glance

Property Individual Observation \(X\) Sample Mean \(\bar{X}\)
Mean \(\mu\) \(\mu\)
Standard Deviation \(\sigma\) \(\sigma / \sqrt{n}\)
Shape Same as population Approximately Normal (large \(n\))
Spread Wider Narrower (decreases with \(n\))
Used for Describing individuals Inference about \(\mu\)

Question 1 of 5

A population has \(\mu = 50\) and \(\sigma = 10\). For a random sample of size \(n = 25\), what is the standard error of \(\bar{X}\)?

Question 2 of 5

The Central Limit Theorem states that the sampling distribution of \(\bar{X}\) is approximately Normal when:

Question 3 of 5

A population is heavily right-skewed with \(\mu = 100\) and \(\sigma = 30\). A sample of \(n = 36\) is taken. What is \(P(\bar{X} < 95)\)?

Question 4 of 5

As sample size \(n\) increases, what happens to the standard error of \(\bar{X}\)?

Question 5 of 5

Which statement best describes the difference between the CLT and the Law of Large Numbers?