BUSS1020 Chap.6 Sampling Distributions and the CLT
Sampling Distributions and the CLT
Sampling distributions connect probability with inference. This chapter separates the population distribution, one observed sample and the distribution of a statistic across repeated samples. It derives the centre and standard error of sample means and proportions, explains how the central limit theorem supports normal approximations, and applies the finite population correction when sampling without replacement.
The worked examples emphasise that an individual standard deviation and a statistic's standard error answer different questions. Increasing sample size narrows the distribution of an average without making individual observations less variable.
The proportion section checks expected successes and failures before using a normal curve, and the finite-population section explains why observing a large share without replacement reduces remaining uncertainty. Clustered data provide the counterexample: a large nominal sample can contain much less independent information than its row count suggests.
Each probability conclusion identifies repeated comparable samples as its subject, stopping a tail area for a statistic from being misreported as a probability about one individual. Direction and denominator checks connect the diagram directly to the calculation. Rebuild that distinction whenever a question shifts from one observation to an average or a proportion.
What this chapter covers
- 01
Population, sample and sampling distributions
- 02
Sampling distribution of the mean
- 03
Standard error and sample size
- 04
Central limit theorem
- 05
Sampling distribution of a proportion
- 06
Finite population correction
Worked example · free
Find a tail probability for a sample mean
- 1Identify the random quantity as a sample mean based on 36 observations.
- 1Compute standard error as 12 divided by the square root of 36, giving 2 minutes.
- 1Standardise 51 as z equal to 51 minus 48 divided by 2, giving 1.5.
- 1Take the upper standard-normal area, approximately 0.0668.
- 1Interpret the probability across comparable random samples, not individual observations.
Key terms
- Sampling distribution
- The probability distribution of a statistic across repeated possible samples of fixed size.
- Standard error
- The standard deviation of a statistic across repeated samples under the model.
- Central limit theorem
- The result that sample-mean distributions approach normality as sample size grows under suitable conditions.
- Sample proportion
- The observed number of successes divided by sample size.
- Finite population correction
- A factor reducing standard error when sampling a substantial fraction without replacement.
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 of a statistic across repeated samples. A sample mean typically uses the population or sample standard deviation divided by the square root of sample size.
Does the central limit theorem make the population normal?
No. It concerns the sampling distribution of the sample mean, not the shape of individual population values. Under suitable conditions, means become approximately normal as sample size grows even when the population itself is not normal.
When should I use the finite population correction?
Use it when sampling without replacement from a finite population and the sample fraction is large enough to matter. The factor is below one because observing a substantial share leaves less unknown population variation.
Exam move
Draw three labelled sketches for every problem: population values, one sample and repeated sample statistics. Write the centre and spread beneath each. Practise changing sample size and predicting the standard-error direction before calculation. For proportions, check expected successes and failures before using a normal approximation.
For finite populations, write N and n and explain why sampling without replacement reduces uncertainty. Use short verbal drills to distinguish an individual probability from a probability about a sample mean or sample proportion. Make a denominator table with σ for one individual, σ divided by square root n for a sample mean, and the binomial proportion expression for p-hat.
Before calculating a tail, state whether averaging should make the distribution narrower than the population. Change n by factors of four and nine and predict the standard-error change using the square-root law.
For the central limit theorem, separate two cases: an exactly normal sample-mean distribution from a normal population, and an approximate result for growing samples from a non-normal population under suitable conditions. Do not use a memorised sample-size cutoff to erase extreme skew or dependence. When sampling a large fraction without replacement, compute the correction and explain why it must be below one.
For clustered users or repeated observations, identify the shared source of dependence and why the nominal sample size overstates information. Finish by interpreting the probability over repeated comparable samples and listing at least two explanations for an unusually extreme statistic besides a true process change.
Working through Sampling Distributions and the CLT in BUSS1020? Sia is AskSia’s AI Statistics tutor — ask any BUSS1020 Sampling Distributions and the CLT question and get a clear, step-by-step explanation grounded in how BUSS1020 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.