Sampling Distributions & CLT - College Statistics

The sample mean ( $\bar{x}$ ) is often described as an 'unbiased estimator' of the population mean ( $\mu$ ). What is the precise statistical meaning of this property?
Answer: The mean of the theoretical sampling distribution of $\bar{x}$ , taken over all possible samples of a given size, is equal to $\mu$ . Unbiasedness of an estimator means that the expected value of the estimator is equal to the population parameter it is estimating. For the sample mean, this translates to the idea that the mean of its sampling distribution is equal to the population mean. It does not mean any particular sample mean will equal the population mean (Choice A), nor does it describe convergence as $n$ increases (Choice B, which is the Law of Large Numbers). It also doesn't guarantee symmetry (Choice D), although the distribution may be symmetric.
A simulation is conducted by taking 5,000 separate random samples of size $n=36$ from a population with a mean $\mu=100$ and standard deviation $\sigma=18$ . The mean of each of the 5,000 samples is calculated and stored. If a histogram is created from these 5,000 sample means, what would be the best estimate for the standard deviation of the values shown in this histogram?
Answer: 3 The histogram described is an empirical model of the sampling distribution of the sample mean ( $\bar{x}$ ). The standard deviation of this distribution is the standard error, $\sigma_{\bar{x}}$ . The formula is $\sigma_{\bar{x}} = \sigma / \sqrt{n}$ . Using the given values, $\sigma_{\bar{x}} = 18 / \sqrt{36} = 18 / 6 = 3$ . The number of samples in the simulation (5,000) is irrelevant to the calculation of the theoretical standard error.
The service time at a bank teller is a random variable with a mean of 3.5 minutes and a standard deviation of 1.2 minutes. Assume the service times for different customers are independent. A manager observes a random sample of 64 customers. According to the Central Limit Theorem, the distribution of the total service time for these 64 customers can be approximated by which of the following?
Answer: A normal distribution with a mean of 224 minutes and a standard deviation of 9.6 minutes. When you encounter problems about the sum or total of independent random variables, you need to apply the properties of sums along with the Central Limit Theorem. The key insight is that we're looking at the distribution of the total service time, not the sample mean. For the sum of n independent random variables, each with mean μ and standard deviation σ, the sum has mean nμ and standard deviation σ√n. Here, with 64 customers each having mean service time 3.5 minutes, the total has mean 64 × 3.5 = 224 minutes. For the standard deviation of the total: σ√n = 1.2√64 = 1.2 × 8 = 9.6 minutes. The Central Limit Theorem tells us this sum will be approximately normally distributed regardless of the original distribution's shape. Choice A incorrectly uses the standard error for a sample mean (σ/√n = 1.2/8 = 0.15) rather than the standard deviation of a sum, and uses the wrong mean. Choice B has the correct mean but uses the original standard deviation (1.2) instead of the sum's standard deviation. Choice C correctly identifies the mean as 224 but fails to recognize that the Central Limit Theorem guarantees the sum will be normally distributed, not maintain the original shape. Therefore, D is correct: a normal distribution with mean 224 minutes and standard deviation 9.6 minutes. Remember: for sums of random variables, multiply the mean by n and multiply the standard deviation by √n. Don't confuse this with sample means, where you divide the standard deviation by √n.
A population's distribution is strongly bimodal, with peaks at 10 and 90. A researcher draws a random sample of size $n=4$ from this population and computes the sample mean. If this process were repeated many times, what would be the most likely shape of the resulting distribution of sample means?
Answer: Unimodal and roughly symmetric, but likely not perfectly normal, centered near 50. With a very small sample size like $n=4$ , the Central Limit Theorem will not have a strong enough effect to make the sampling distribution fully normal, especially from a strongly bimodal population. However, the process of averaging will still tend to pull the means away from the extremes and toward the center. For example, a sample might contain values from both modes, resulting in a mean near 50. The resulting distribution will therefore lose the strong bimodality of the parent population and tend towards a central peak, but it will not yet be bell-shaped. It will be more peaked and less bimodal than the population.
The formula for the standard error of a sample mean, $\sigma/\sqrt{n}$ , relies on the assumption that individual observations are independent. This assumption is technically violated when sampling without replacement. Why is it still considered acceptable to use this formula when the sample size $n$ is less than 10% of the population size $N$ ?
Answer: Because removing a small fraction of the population does not meaningfully change the remaining probabilities, so the observations are 'nearly' independent. When you encounter questions about sampling without replacement, you need to understand why we can still use independence-based formulas despite technically violating the independence assumption. The key insight is that independence in sampling is about whether selecting one item affects the probability of selecting another. When you sample without replacement from a large population, removing a few observations barely changes the composition of the remaining population. If you remove 50 people from a population of 10,000, the characteristics of the remaining 9,950 people are virtually unchanged. This makes successive draws "nearly independent" even though they're technically dependent. The correct answer is D because it captures this practical reality. When your sample is less than 10% of the population, the finite population correction factor $\sqrt{\frac{N-n}{N-1}}$ is so close to 1 that it's negligible, meaning the standard error formula $\sigma/\sqrt{n}$ remains highly accurate. Answer A is wrong because normality of the sampling distribution isn't the issue here—we're discussing independence, not distribution shape. Answer B incorrectly suggests bias is the concern; the sample mean is unbiased regardless of the 10% condition. Answer C confuses the 10% rule with sample size requirements for the Central Limit Theorem—these are separate concepts. Remember this pattern: The 10% rule is about maintaining "practical independence" when sampling without replacement. It's not about normality, bias, or CLT requirements—it's purely about keeping dependence effects negligible.
The number of typos per page in a book manuscript follows a discrete, right-skewed distribution with a mean of 2.5 typos and a standard deviation of 1.5 typos. An editor randomly selects 50 pages and calculates the average number of typos per page, $\bar{x}$ . Which statement best describes the sampling distribution of $\bar{x}$ ?
Answer: It is approximately normal with a mean of 2.5 and a standard deviation of $1.5/\sqrt{50}$ . The Central Limit Theorem applies to samples from any population distribution, including discrete and skewed ones, provided the sample size is sufficiently large. Here, $n=50$ is large enough. The CLT states that the sampling distribution of $\bar{x}$ will be approximately normal. Its mean, $\mu_{\bar{x}}$ , is equal to the population mean, $\mu = 2.5$ . Its standard deviation (the standard error), $\sigma_{\bar{x}}$ , is the population standard deviation divided by the square root of the sample size, $\sigma/\sqrt{n} = 1.5/\sqrt{50}$ .
A researcher takes a single random sample of $n=40$ exam scores from a large population of scores, which is known to be left-skewed. Which statement correctly distinguishes between the 'distribution of the sample' and the 'sampling distribution of the sample mean' in this context?
Answer: The distribution of the sample will likely be left-skewed, while the sampling distribution of the sample mean will be approximately normal. These are two distinct concepts. The 'distribution of the sample' refers to the distribution of the 40 actual data points collected. As it's a random sample from a left-skewed population, it is also expected to be left-skewed. The 'sampling distribution of the sample mean' is a theoretical distribution of all possible sample means of size 40. Since the sample size $n=40$ is large, the Central Limit Theorem applies, and this theoretical distribution will be approximately normal.
A quality control specialist is examining the standard deviation of the sampling distribution of the mean weight of a product. For a sample of size $n$ , the standard error is $\sigma_{\bar{x}}$ . If the specialist wants to reduce this standard error to one-third of its original value (i.e., to $\frac{1}{3}\sigma_{\bar{x}}$ ), what new sample size must be taken?
Answer: $9n$ The standard error of the mean is given by the formula $\sigma_{\bar{x}} = \frac{\sigma}{\sqrt{n}}$ , where $\sigma$ is the population standard deviation and $n$ is the sample size. To reduce the standard error to $\frac{1}{3}$ of its original value, we need to solve for the new sample size, $n'$ : $\frac{\sigma}{\sqrt{n'}} = \frac{1}{3} \left( \frac{\sigma}{\sqrt{n}} \right)$ . This simplifies to $\sqrt{n'} = 3\sqrt{n}$ . Squaring both sides gives $n' = 9n$ . Therefore, the sample size must be multiplied by 9.
A political analyst calculates a 95% confidence interval for voter turnout and reports that the standard error of the sample proportion was 0.02. What does this value of 0.02 represent?
Answer: The approximate average distance between the sample proportion and the true population proportion over many hypothetical samples. The standard error of a statistic (in this case, the sample proportion) is the standard deviation of its sampling distribution. Conceptually, it quantifies the typical amount of error, or the average distance, between the values of the statistic from all possible samples of a given size and the actual population parameter. It is not the standard deviation of a single sample's data (A), nor is it the full margin of error (C), which is typically $z^* \times SE$ . It is also not a probability like alpha (D).
A machine fills bottles with a mean of 500 ml and a standard deviation of 2 ml. The distribution of fill volumes is symmetric and unimodal. An inspector takes a random sample of 16 bottles. What is the mean of the sampling distribution of the sample mean fill volume ( $\bar{x}$ )?
Answer: 500 ml A fundamental property of sampling distributions is that the mean of the sampling distribution of the sample mean ( $\mu_{\bar{x}}$ ) is always equal to the mean of the population ( $\mu$ ) from which the samples are drawn. This is true regardless of the sample size or the shape of the population distribution. Given the population mean is 500 ml, the mean of the sampling distribution of $\bar{x}$ is also 500 ml.
A company's employee performance scores are measured on a 1-to-5 integer scale. The true distribution of scores for all employees is: 10% score 1, 15% score 2, 20% score 3, 40% score 4, and 15% score 5. A manager takes a random sample of 120 employees. Which statement most accurately describes the sampling distribution of the average performance score, $\bar{x}$ ?
Answer: It will be approximately normal with a mean of 3.35. First, find the population mean ( $\mu$ ). $\mu = (1)(0.10) + (2)(0.15) + (3)(0.20) + (4)(0.40) + (5)(0.15) = 0.10 + 0.30 + 0.60 + 1.60 + 0.75 = 3.35$ . The mean of the sampling distribution of $\bar{x}$ is equal to $\mu$ . Since the sample size $n=120$ is large, the Central Limit Theorem applies, and the shape of the sampling distribution will be approximately normal, regardless of the population's discrete, non-symmetric shape. Therefore, the sampling distribution is approximately normal with a mean of 3.35.
Two different populations of component lifetimes are being studied. Population X has a normal distribution with mean $\mu_X = 1000$ hours and standard deviation $\sigma_X = 100$ hours. Population Y has an exponential distribution (which is highly right-skewed) with mean $\mu_Y = 1000$ hours and standard deviation $\sigma_Y = 1000$ hours. For which of the following sample sizes $n$ would the sampling distribution of $\bar{x}$ from Population X be most similar in shape to the sampling distribution of $\bar{y}$ from Population Y?
Answer: $n=50$ for both The sampling distribution from Population X will be exactly normal for any sample size because the population is normal. The sampling distribution from Population Y will become approximately normal for a large sample size due to the CLT. For the two sampling distributions to be most similar in shape (i.e., both approximately normal), we need a large enough sample size for the CLT to work its effect on Population Y. With $n=5$ , the sampling distribution from Y would still be quite skewed. With $n=50$ for both, the sampling distribution from X is exactly normal, and the sampling distribution from Y is approximately normal, making them very similar in shape.
A researcher wants to estimate the proportion, $p$ , of people who own a specific brand of smartphone. They plan to survey a random sample of $n=120$ people. In which of the following scenarios would the normal approximation to the sampling distribution of the sample proportion, $\hat{p}$ , be the LEAST appropriate to use?
Answer: The true population proportion is believed to be $p=0.95$ . The normal approximation for the sampling distribution of a sample proportion is considered appropriate when the Large Counts Condition is met: $np \ge 10$ and $n(1-p) \ge 10$ . We check this for each choice with $n=120$ . A) $n(1-p) = 120(1-0.95) = 120(0.05) = 6$ . Since $6 < 10$ , the condition is not met. B) $np = 120(0.50) = 60$ and $n(1-p) = 120(0.50) = 60$ . Both are $\ge 10$ . C) $np = 120(0.20) = 24$ and $n(1-p) = 120(0.80) = 96$ . Both are $\ge 10$ . The condition is violated when the proportion $p$ is very close to 0 or 1, making choice A the least appropriate scenario.
A statistics student claims, 'For the sampling distribution of the sample mean to be approximately normal, the sample size must be at least 30.' Which of the following scenarios presents a valid reason why this statement is not universally correct?
Answer: If the population distribution is itself normal, the sampling distribution of the mean is exactly normal for any sample size. The rule of thumb that $n \ge 30$ is a guideline for applying the Central Limit Theorem when the population distribution is not normal. However, if the population from which samples are drawn is already normally distributed, the sampling distribution of the sample mean will be exactly normal for any sample size, $n \ge 1$ . This scenario demonstrates that $n \ge 30$ is a sufficient condition in many cases but not a universally necessary one.
As the sample size $n$ taken from a population with mean $\mu$ and finite standard deviation $\sigma$ increases, which of the following correctly describes the changes in the sampling distribution of the sample mean $\bar{x}$ ?
Answer: The mean of the sampling distribution remains centered at $\mu$ , and its standard error decreases. When you encounter questions about sampling distributions, you're dealing with one of the most fundamental concepts in statistics: how sample means behave as sample size changes. This is governed by the Central Limit Theorem and basic properties of sampling distributions. As sample size $n$ increases, two key things happen to the sampling distribution of $\bar{x}$ . First, the mean of the sampling distribution always equals the population mean $\mu$ , regardless of sample size – this is an unbiased estimator property. Second, the standard error (standard deviation of the sampling distribution) equals $\frac{\sigma}{\sqrt{n}}$ , which clearly decreases as $n$ increases. Answer D correctly captures both unchanging properties: the mean stays at $\mu$ and the standard error decreases with larger samples. Answer A is wrong because the mean of the sampling distribution never changes – it's always $\mu$ . Answer B incorrectly states that standard error increases; it actually decreases as $n$ grows. The first part about approaching normality is true due to the Central Limit Theorem, but the second part makes this answer incorrect. Answer C is backwards about shape – the sampling distribution approaches normality (not the parent population's shape) and becomes more concentrated around $\mu$ as $n$ increases. Remember this pattern: larger samples always give you more precise estimates (smaller standard error) while maintaining the same expected value. The formula $SE = \frac{\sigma}{\sqrt{n}}$ is crucial – memorize it and understand that it decreases as $n$ increases.
Let Population A be a normal distribution with mean 50 and standard deviation 12. Let Population B be a uniform distribution on the interval [0, 100], which has a mean of 50. A researcher creates sampling distribution $S_A$ by taking samples of size $n=4$ from Population A, and sampling distribution $S_B$ by taking samples of size $n=36$ from Population B. Which statement best compares the characteristics of $S_A$ and $S_B$ ?
Answer: $S_A$ will be exactly normal, while $S_B$ will be approximately normal, and both will have the same mean. The mean of any sampling distribution of the mean is equal to the population mean, so both $S_A$ and $S_B$ will have a mean of 50. Since Population A is normal, the sampling distribution $S_A$ will be exactly normal, even with a small sample size of $n=4$ . Population B is not normal (it's uniform), but since the sample size $n=36$ is large, the Central Limit Theorem states that the sampling distribution $S_B$ will be approximately normal. Therefore, $S_A$ is exactly normal, $S_B$ is approximately normal, and they share the same mean.
Which of the following statements provides the most accurate distinction between the Law of Large Numbers (LLN) and the Central Limit Theorem (CLT)?
Answer: The LLN states that the sample mean converges to the population mean as $n$ increases, while the CLT describes the shape of the sampling distribution for a fixed, large $n$ . The two theorems describe different phenomena related to sampling. The Law of Large Numbers is about convergence: as the sample size $n$ grows, the sample mean $\bar{x}$ gets arbitrarily close to the population mean $\mu$ . The Central Limit Theorem is about the shape of the sampling distribution: for a large (but fixed) sample size $n$ , the theoretical distribution of all possible sample means is approximately normal.
A population of annual household incomes is known to be strongly right-skewed with a mean of $70,000 and a standard deviation of $50,000. A researcher plans to take a random sample of size $n=400$ and calculate the sample mean income, $\bar{x}$ . Which of the following statements gives the most accurate description of the sampling distribution of $\bar{x}$ ?
Answer: The distribution will be approximately normal with a mean of $70,000. The Central Limit Theorem (CLT) states that for a sufficiently large sample size, the sampling distribution of the sample mean ( $\bar{x}$ ) will be approximately normal, regardless of the shape of the population distribution. A sample size of $n=400$ is considered large. The mean of this sampling distribution ( $\mu_{\bar{x}}$ ) is equal to the population mean ( $\mu$ ), which is $70,000. Therefore, the sampling distribution of $\bar{x}$ will be approximately normal with a mean of $70,000.
Standard formulas for standard error, such as $\sigma/\sqrt{n}$ , assume that samples are drawn with replacement or from an infinitely large population. When sampling without replacement from a finite population, these formulas are an approximation. In which of the following scenarios would the standard formula provide the least accurate estimate of the true standard error?
Answer: A sample of $n=400$ is drawn from a population of $N=1,000$ . The accuracy of the standard formula depends on the ratio of sample size to population size ( $n/N$ ). The approximation is poor when this ratio is large. A common rule of thumb is that the approximation is acceptable if $n/N \le 0.05$ or $n/N \le 0.10$ . Let's calculate the ratio for each option: A) $50/500,000 = 0.0001$ B) $100/2,000 = 0.05$ C) $400/1,000 = 0.40$ D) $1,000/10,000,000 = 0.0001$ The ratio is largest in choice C, where the sample is 40% of the population. In this case, the standard formula would significantly overestimate the true standard error, making it the least accurate.
The graphic shows the probability distribution for a discrete random variable, representing a population. Four different empirical sampling distributions (A, B, C, D) were generated by taking 1,000 samples of a certain size $n$ from this population and plotting the sample means. Based on the Central Limit Theorem, which of the distributions most likely corresponds to the largest sample size $n$ ?
Answer: Distribution C The Central Limit Theorem states that as the sample size $n$ increases, the sampling distribution of the sample mean becomes more normally distributed (more symmetric and bell-shaped) and its variance decreases. The population distribution is clearly skewed to the right. Distribution D is highly skewed, suggesting a very small $n$ (likely $n=1$ or $n=2$ ). Distributions A and B show decreasing skewness and variance. Distribution C is the most symmetric, bell-shaped, and has the smallest spread (is the narrowest), which are all characteristics of a sampling distribution generated with the largest sample size.

Start this quiz