QBUS5001 Chap.6 Sampling Distributions and the CLT
Sampling Distributions and the CLT
This chapter treats a sample mean or proportion as a random variable before data are collected. Repeated samples create a sampling distribution whose centre and standard error determine estimator precision. The Central Limit Theorem supports an approximate Normal shape for many large-sample means, while the Law of Large Numbers explains convergence. Neither result repairs selection bias or a weak sampling frame.
The population distribution describes individual values, one realised sample has its own observed shape, and the sampling distribution describes a statistic across hypothetical repeated samples. Keeping those three layers separate prevents many inference errors. For a sample mean, the centre equals the population mean and the standard error scales the population standard deviation by the square root of sample size.
For a sample proportion, the centre equals the population proportion and its standard error reflects both the proportion and sample size. Increasing sample size narrows either sampling distribution without changing its target. The Central Limit Theorem concerns the shape of a standardised statistic under appropriate conditions; it does not make the raw population Normal.
The Law of Large Numbers concerns convergence, not a guarantee that one large convenience sample is representative.
What this chapter covers
- 01
Statistics as random variables across repeated samples
- 02
Unbiased centres for sample means and proportions
- 03
Standard error and the square-root sample-size rule
- 04
Central Limit Theorem for sample means
- 05
Sampling distribution of a sample proportion
- 06
Precision versus representativeness
Worked example · free
Approximate a sample-mean tail
- +1Identify the random object as the mean of 100 invoice values.
- +1Compute standard error 30 divided by square root 100 equals 3.
- +1Standardise: 86 minus 80 divided by 3 equals 2.
- +1Use the right Normal tail at 2 to obtain approximately 0.0228.
- +1Interpret the probability for the sample average, not one invoice.
Key terms
- Sampling Distribution
- The probability distribution of a statistic across repeated samples from the same mechanism.
- Standard Error
- The standard deviation of a statistic across repeated samples.
- Unbiased Estimator
- An estimator whose sampling distribution is centred on the population parameter.
- Central Limit Theorem
- A result supporting an approximately Normal distribution for many standardised sample means when sample size is sufficiently large.
Sampling Distributions and the CLT FAQ
What is the difference between standard deviation and standard error?
Standard deviation describes variability among individual observations. Standard error describes variability among repeated values of a statistic, such as sample means. For a mean under the basic model, standard error shrinks with the square root of sample size.
Does the Central Limit Theorem make raw data Normal?
No. It concerns the sampling distribution of a standardised sum or mean under suitable conditions. The population data can remain skewed. A large sample may make the mean approximately Normal without changing the shape of individual observations.
Can a very large sample still be biased?
Yes. A large sample can produce a very small random standard error while systematically excluding part of the target population. Precision describes repeated sampling from the same mechanism; representativeness depends on how that mechanism covers the population.
Exam move
Draw three layers for every question: population distribution, one sample statistic and the distribution of that statistic across repeated samples. Label the correct centre and standard error. Practise converting a raw threshold into standard-error units for both means and proportions. Finish by writing the sampling-frame assumption, because a narrow standard error does not establish representativeness.
Use a small simulation table to imagine five possible samples from the same population and record the mean of each. The variation among those means is the object that the standard error measures. Double the sample size and predict the narrowing factor before calculating it; the square-root relationship matters more than memorising one numeric result.
Pair a mean problem with a proportion problem so the shared logic and different standard-error formulas remain distinct. For every Central Limit Theorem statement, name what becomes approximately Normal and the conditions supporting that approximation.
Close each exercise by separating precision from validity: a statistic can have a very small standard error and still be biased if the sampling mechanism systematically misses part of the population. Link each standard error to a hypothetical repetition statement: if the same sampling mechanism were repeated, how would this statistic move?
That question keeps the denominator attached to the sample size and prevents the spread of individual observations from being substituted for estimator uncertainty. It also prepares the interpretation needed for intervals and tests in the following chapters.
Working through Sampling Distributions and the CLT in QBUS5001? Sia is AskSia’s AI Statistics tutor — ask any QBUS5001 Sampling Distributions and the CLT question and get a clear, step-by-step explanation grounded in how QBUS5001 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.