Sample means and sampling distributions (SACE Stage 2 Mathematical Methods)

Syllabus dot point

“The sampling distribution of the sample mean is approximately normal, centred on the population mean, with a standard deviation that shrinks as the sample size grows.”

8 min readUpdated 2 Jun 2026

Quick answer

If you take many samples of size $n$ and record each sample mean $\bar{x}$ , those means form a sampling distribution. It is centred on the population mean ( $\mu_{\bar X}=\mu$ ) with standard deviation (the standard error) $\dfrac{\sigma}{\sqrt{n}}$ , and by the Central Limit Theorem it is approximately normal for large $n$ .

Jump to a section

What this dot point is asking
Mean and standard error
The Central Limit Theorem
Sums versus means
Why averaging reduces spread
Common errors
Why it matters
Exam-style questions

What this dot point is asking

A single sample gives one estimate of the population mean, but if you took another sample you would get a slightly different $\bar{x}$ . The sampling distribution of the sample mean describes how these sample means vary from sample to sample.

Sampling distribution of the mean

The sampling distribution of the sample mean $\bar{X}$ is the probability distribution of the values of $\bar{x}$ obtained from all possible samples of a fixed size $n$ drawn from the population.

Mean and standard error

The sampling distribution has a smaller spread than the population, because averaging cancels out extreme individual values.

Mean and standard error of the sample mean

\mu_{\bar X}=\mu, \qquad \sigma_{\bar X}=\frac{\sigma}{\sqrt{n}}.

The quantity

\dfrac{\sigma}{\sqrt{n}}

is called the standard error of the mean.

The standard error shrinks as $n$ grows, but only with $\sqrt{n}$ - to halve the spread you must take four times as many observations.

Standard error and sample size

A population has $\mu=50$ and $\sigma=12$ . This example finds the mean and standard error of the sampling distribution of $\bar{X}$ for samples of size $n=36$ .

Step 1: state the mean of the sampling distribution

The mean of the sampling distribution of $\bar{X}$ always equals the population mean, regardless of sample size. Averaging does not introduce any systematic shift, so:

\mu_{\bar X}=\mu=50.

Step 2: compute the standard error

The standard error measures how much sample means vary from sample to sample. It equals the population standard deviation divided by the square root of the sample size, so larger samples produce less variable means:

\sigma_{\bar X}=\frac{\sigma}{\sqrt{n}}=\frac{12}{\sqrt{36}}=\frac{12}{6}=2.

Final answer: the sampling distribution has mean $50$ and standard error $2$ , far tighter than the population spread of $12$ .

The Central Limit Theorem

Central Limit Theorem

For a sufficiently large sample size (a common rule of thumb is $n\ge 30$ ), the sampling distribution of $\bar{X}$ is approximately normal, regardless of the shape of the original population. If the population is itself normal, $\bar{X}$ is exactly normal for any $n$ .

This is what makes statistical inference possible: even with a skewed population, you can use normal-distribution methods on the sample mean.

Probability for a sample mean

Using the same population ( $\mu=50$ , $\sigma=12$ , $n=36$ , giving $\sigma_{\bar X}=2$ ), this example finds the probability that a sample mean exceeds $53$ .

Step 1: standardise using the standard error, not $\sigma$

When working with a sample mean rather than a single observation, the denominator of the z-score formula is the standard error $\sigma/\sqrt{n}$ , not the population standard deviation $\sigma$ . Using $\sigma$ alone would treat the sample mean as if it were one individual value and drastically overstate the spread:

z=\frac{\bar x-\mu}{\sigma/\sqrt{n}}=\frac{53-50}{2}=1.5.

Step 2: find the tail probability

The event $P(\bar X>53)$ is a right-tail area. Calculators return left-tail areas by default, so we subtract from $1$ :

P(\bar X>53)=P(Z>1.5)=1-P(Z<1.5)=1-0.9332=0.0668.

Final answer: there is approximately a $6.7\%$ chance that a sample of $36$ observations has a mean exceeding $53$ .

Sums versus means

SACE questions phrase the same idea two ways: the sample mean $\bar X$ and the sample sum $S_n=X_1+X_2+\dots+X_n$ . For the sum, the mean is $n\mu$ and the standard deviation is $\sqrt{n}\,\sigma$ , because both the means and the variances of independent observations add. For the mean, dividing the sum by $n$ gives mean $\mu$ and standard deviation $\dfrac{\sigma}{\sqrt{n}}$ . Notice the contrast: the spread of a sum grows with $\sqrt{n}$ while the spread of a mean shrinks with $\sqrt{n}$ . Reading whether the question is about a total or an average decides which result you use.

Why averaging reduces spread

It is worth understanding why the standard error shrinks. In any sample, some observations land above the population mean and some below; when you average them, the high and low values partly cancel. The larger the sample, the more complete this cancellation, so sample means cluster ever more tightly around $\mu$ . The $\sqrt{n}$ rather than $n$ in the denominator reflects that this cancellation is partial - randomness does not vanish, it just averages down at the rate of the square root of the sample size, which is the law of diminishing returns behind every sample-size calculation.

Common errors

Using σ instead of the standard error σ/√n

When standardising a sample mean, you must divide by the standard error $\sigma/\sqrt{n}$ , not the population standard deviation $\sigma$ . Using $\sigma$ alone treats $\bar X$ as if it were a single observation and badly overstates the spread. The $\sqrt{n}$ in the denominator is the whole point - it is why sample means are more reliable than individual values.

Distinguish "a value" from "a mean"

Read the question carefully: $P(X>53)$ for one individual uses $\sigma$ , but $P(\bar X>53)$ for a sample mean uses $\sigma/\sqrt{n}$ . Underline whether the quantity is a single observation or an average before you standardise.

Why it matters

The sampling distribution is the conceptual core of Topic 6. The standard error $\sigma/\sqrt{n}$ and the Central Limit Theorem are exactly what allow you to build confidence intervals for a population mean in the next dot points, and they explain why larger samples give more precise estimates.

Exam-style questions

Questions in the style of SACE Board exam questions on this dot point, each with a worked answer. They are written by ExamExplained unless tagged "Past paper"; the year shows the paper a question is modelled on.

SACE 2023-style2 marks

Calculator-assumed. Daily milk production per cow has mean

21.9

litres and standard deviation

3.26

litres. Let

S_{90}

be the total daily production of a random sample of 90 cows. Show that

S_{90}

has mean

1971

litres and standard deviation

30.93

litres (to four significant figures).

Show worked answer →

For a sum of $n$ independent observations, the means add and the variances add (so the standard deviation is multiplied by $\sqrt{n}$ ).

Mean of $S_{90} = n\mu = 90 \times 21.9 = 1971$ litres.

Standard deviation of $S_{90} = \sqrt{n}\,\sigma = \sqrt{90} \times 3.26 = 9.4868 \times 3.26 \approx 30.93$ litres.

Marks: one for the mean, one for the standard deviation using $\sqrt{90}$ (not $90$ ). Only variances add directly, so the standard deviation scales with $\sqrt{n}$ .

SACE 2022-style2 marks

Calculator-assumed. The number of shares of a meme has mean

131.25

and standard deviation

221

. Let

S_{75}

be the sum of the shares of a random sample of 75 memes. Show that

S_{75}

has mean

9844

shares and standard deviation

1914

shares.

Show worked answer →

For the sum of $n$ independent variables, multiply the mean by $n$ and the standard deviation by $\sqrt{n}$ .

Mean of $S_{75} = 75 \times 131.25 = 9843.75 \approx 9844$ shares.

Standard deviation of $S_{75} = \sqrt{75} \times 221 = 8.6603 \times 221 \approx 1914$ shares.

Marks: one for the mean, one for the standard deviation using $\sqrt{75}$ . By the Central Limit Theorem $S_{75}$ is also approximately normal.

SACE 2021-style3 marks

Calculator-assumed. A population has mean

\mu = 80

and standard deviation

\sigma = 15

. A random sample of size

n = 25

is taken. (a) State the mean and standard error of

\bar{X}

. (b) Find

\Pr(\bar{X} < 76)

Show worked answer →

(a) $\mu_{\bar X} = 80$ and $\sigma_{\bar X} = \dfrac{\sigma}{\sqrt{n}} = \dfrac{15}{\sqrt{25}} = \dfrac{15}{5} = 3$ . (1 mark)

(b) Standardise the sample mean using the standard error: $z = \dfrac{76 - 80}{3} = -1.333$ . Then $\Pr(\bar X < 76) = \Pr(Z < -1.333) \approx 0.0912$ . (2 marks)

Marks reward using the standard error $\sigma/\sqrt{n}$ , not $\sigma$ , when standardising a sample mean.

Practise this

Sources & how we know this

Stage 2 Mathematical Methods subject outline (2025 onwards) — SACE Board
Stage 2 Mathematical Methods: external assessment — SACE Board

What this dot point is asking

Mean and standard error

Step 1: state the mean of the sampling distribution

Step 2: compute the standard error

The Central Limit Theorem

Step 1: standardise using the standard error, not σ\sigmaσ

Step 2: find the tail probability

Sums versus means

Why averaging reduces spread

Common errors

Why it matters

Exam-style questions

Practise this

Sources & how we know this

Step 1: standardise using the standard error, not $\sigma$