Quick Summary
This guide will empower you to understand and apply the Central Limit Theorem (CLT), one of the most foundational concepts in statistics. You will learn to describe the shape, center, and spread of the sampling distribution of a sample mean (x̄). This involves checking critical conditions to determine when this distribution can be modeled as Approximately Normal, a skill essential for performing inference on population means.
Key Concepts
The Central Limit Theorem is a powerful idea that allows us to make predictions about a population mean using a single sample mean. To understand it, we must first understand the sampling distribution of the sample mean (x̄).
The Idea: Imagine you have a population (e.g., all adult males in the U.S.). You could take a random sample of size n (say, 50 men) and calculate their average height, x̄. You could then take a different random sample of 50 men and get a slightly different x̄. If you did this thousands of times, the distribution of all those different x̄ values is the sampling distribution of the sample mean.
The Central Limit Theorem (CLT): The theorem tells us what this sampling distribution will look like. It states:
For a population with any shape, if the sample size n is sufficiently large (n \ge 30), the sampling distribution of the sample mean x̄ will be Approximately Normal.
The Two Paths to Normality: There are two ways we can know that the sampling distribution of x̄ is Approximately Normal:
The Population is Normal: If the original population distribution is stated to be Normal, then the sampling distribution of x̄ will also be Normal, regardless of the sample size n. It could be n=5 or n=500; the result is a Normal sampling distribution.
The Central Limit Theorem Applies: If the population distribution is skewed, uniform, or its shape is unknown, we must rely on the CLT. As long as our sample size is large enough (n \ge 30), we can say the sampling distribution of x̄ is Approximately Normal.
[Image: A series of three graphs. The first graph shows a population distribution that is heavily skewed to the right. The second graph shows the sampling distribution of x̄ for n=5, which is still somewhat skewed right. The third graph shows the sampling distribution of x̄ for n=30, which is symmetric, unimodal, and bell-shaped, labeled "Approximately Normal by CLT".]
Describing the Sampling Distribution of x̄: Just like any distribution, we describe it using its Shape, Center, and Spread.
Shape: Is it Approximately Normal? Check the conditions above (Population Normal OR n \ge 30).
Center (Mean): The mean of the sampling distribution of x̄, denoted μₓ̄, is always equal to the true population mean, μ.
Formula: μₓ̄ = μ
This means our sample means will, on average, center on the true population mean. It's an unbiased estimator.
Spread (Standard Deviation): The standard deviation of the sampling distribution of x̄, denoted σₓ̄, is calculated from the population standard deviation, σ, and the sample size, n. This is also called the standard error of the mean.
Formula: σₓ̄ = σ / √n
Condition: This formula is only accurate if our sample is independent. We check this using the 10% Condition: the sample size n must be no more than 10% of the population size N (i.e., n \le 0.10N). If we are sampling without replacement and this condition isn't met, the formula is not valid.
Notice that as the sample size n increases, the denominator gets larger, making the standard deviation smaller. Larger samples lead to less variability in the sample means, which makes intuitive sense—larger samples give more precise estimates.
Key Vocabulary
Central Limit Theorem (CLT): The theorem stating that the sampling distribution of the sample mean (x̄) will be approximately normal if the sample size is sufficiently large (n \ge 30), regardless of the population's shape.
Sampling Distribution of a Sample Mean (x̄): The theoretical probability distribution of all possible sample means (x̄) of a given sample size n taken from a population.
Population Distribution: The distribution of values of a variable for all individuals in the entire population.
Sample Distribution: The distribution of values of a variable for the individuals in one specific sample. Do not confuse this with the sampling distribution.
Standard Error of the Mean: The standard deviation of the sampling distribution of the sample mean (σₓ̄). It measures the typical distance between a sample mean (x̄) and the population mean (μ).
10% Condition: A rule of thumb for checking the independence of observations when sampling without replacement. It requires that the sample size n is no more than 10% of the population size N.
Calculator Tech (TI-84)
No major new calculator functions are required for understanding the Central Limit Theorem itself. However, when solving probability problems involving a Normal sampling distribution, you will use the function.
mean: Use the mean of the sampling distribution, μₓ̄.
std dev: Use the standard deviation of the sampling distribution, σₓ̄ = σ/√n.
How to Show Work on the FRQ
On Free Response Questions, you must justify your calculations by fully describing the sampling distribution of the sample mean. Follow this four-step process, often called "State, Plan, Do, Conclude."
Template for Calculating a Probability for a Sample Mean (x̄):
1. STATE:
Define the parameter of interest: "Let μ be the true mean [context of the problem]."
State the statistic you are using: "We are given a sample of size n = [value] with a sample mean of x̄ = [value]."
State what you want to find: "We want to find the probability that x̄ is [less than/greater than/between] [value(s)], or P(x̄ > [value])."
2. PLAN (Check Conditions & Describe the Sampling Distribution):
Random: "The problem states the sample was randomly selected." (Or assume it is representative).
10% Condition: "The sample size n = [value] is less than 10% of the total population of [context of population], so we can assume independence." (e.g., "n=50 is less than 10% of all coffee drinkers").
Normal/Large Sample: "Because the [population is stated to be Normal / sample size n = [value] is \ge 30], the Central Limit Theorem applies. Therefore, the sampling distribution of x̄ is Approximately Normal."
Describe the Distribution:
Shape: Approximately Normal (justified above).
Center: The mean of the sampling distribution is μₓ̄ = μ = [value from problem].
Spread: The standard deviation of the sampling distribution is σₓ̄ = σ/√n = [value]/√[n] = [calculated value].
3. DO (Calculations):
Draw a Normal curve, label the mean (μₓ̄), and shade the area of interest.
Calculate the z-score: z = (x̄ - μₓ̄) / σₓ̄ = ([value] - [mean]) / [std dev].
Use the z-score or calculator to find the probability: P(x̄ > [value]) = P(Z > [z-score]) = . Show your inputs.
4. CONCLUDE:
- Answer the question in the context of the problem: "There is a [calculated probability] probability of observing a sample mean of [context] of [less than/greater than] [value] from a random sample of size n."
Practice Problems
Problem 1: The mean time for a barista at a coffee shop to make a latte is 120 seconds with a standard deviation of 18 seconds. The distribution of times is known to be skewed to the right. A random sample of 36 lattes is selected. What is the probability that the average time to make these 36 lattes is less than 115 seconds?
Solution:
First, we will define the parameter and statistic of interest. Let μ be the true mean time to make a latte, which is 120 seconds. We have a random sample of n=36 lattes and want to find the probability that the sample mean time, x̄, is less than 115 seconds, or P(x̄ < 115).
Next, we must check the conditions to describe the sampling distribution of x̄. The problem states the sample was selected randomly. The sample size of n=36 is surely less than 10% of all lattes made at this shop, so the 10% condition is met. Although the population distribution is skewed right, our sample size is n=36, which is \ge 30. Therefore, by the Central Limit Theorem, the sampling distribution of x̄ is Approximately Normal. The mean of this distribution is μₓ̄ = μ = 120 seconds. The standard deviation is σₓ̄ = σ/√n = 18/√36 = 18/6 = 3 seconds.
Now we can perform the calculation. We want to find P(x̄ < 115). The z-score is z = (115 - 120) / 3 = -5 / 3 \approx -1.67. The probability is P(Z < -1.67), which can be found using normalcdf(lower: -1E99, upper: 115, mean: 120, std dev: 3) = 0.0478.
In conclusion, there is approximately a 0.0478 probability of observing a sample mean latte preparation time of less than 115 seconds from a random sample of 36 lattes.
Problem 2: The weights of golden retriever puppies at 8 weeks of age are normally distributed with a mean of 15 pounds and a standard deviation of 1.2 pounds. A breeder takes a random sample of 9 puppies. What is the probability that the average weight of these 9 puppies is between 14.5 and 16 pounds?
Solution:
First, we define our parameter. Let μ be the true mean weight of 8-week-old golden retriever puppies, which is 15 pounds. We have a random sample of n=9 puppies and want to find the probability that the sample mean weight, x̄, is between 14.5 and 16 pounds, or P(14.5 < x̄ < 16).
Next, we check conditions. The sample was randomly selected. The sample size n=9 is less than 10% of all golden retriever puppies, so the 10% condition is met. The sample size is small (n=9), but the problem states that the population of puppy weights is normally distributed. Therefore, the sampling distribution of x̄ is also Normal. The mean of this distribution is μₓ̄ = μ = 15 pounds. The standard deviation is σₓ̄ = σ/√n = 1.2/√9 = 1.2/3 = 0.4 pounds.
Now we perform the calculation for P(14.5 < x̄ < 16). We can use normalcdf(lower: 14.5, upper: 16, mean: 15, std dev: 0.4) = 0.8881. For z-scores, z₁ = (14.5 - 15) / 0.4 = -1.25 and z₂ = (16 - 15) / 0.4 = 2.5. P(-1.25 < Z < 2.5) = 0.8881.
In conclusion, there is an 0.8881 probability that the mean weight of a random sample of 9 puppies will be between 14.5 and 16 pounds.
Common Mistakes to Avoid
Confusing the Sample with the Sampling Distribution: The CLT states that the sampling distribution of the mean becomes normal, NOT that the sample itself becomes normal. If the population is skewed, a single random sample of size 40 will likely still appear skewed, but the distribution of all possible means from samples of size 40 will be approximately normal.
Forgetting to Divide by the Square Root of n: A very common error is to use the population standard deviation (σ) in your z-score calculation instead of the standard deviation of the sampling distribution (σₓ̄ = σ/√n). Always remember to adjust the standard deviation for the sample size.
Applying the CLT with a Small Sample Size: If n < 30, you cannot use the CLT to claim the sampling distribution is approximately normal. You can only proceed if the problem explicitly states the original population is normal.
Misstating the Conclusion of the CLT: Do not say "the sample is normal" or "the population is normal" because of the CLT. The correct phrasing is "the sampling distribution of the sample mean (x̄) is approximately normal." Precision is key.
Forgetting to Check the 10% Condition: While often a given, you must explicitly check the 10% condition (n \le 0.10N) to justify using the formula for the standard deviation of the sampling distribution. On an FRQ, failing to check this can lose you points.