Mathematical Methods · Unit 4 · Topic 4

Sampling and proportions

Choose random samples and spot bias, read displays of real and simulated samples, then treat the sample proportion p^\hat{p} as a random variable with mean pp, standard deviation p(1−p)/n\sqrt{p(1 - p)/n} and an approximately normal distribution for large samples.

  • 7 lessons
  • 1 h 50 min of lessons
  • 11 worked examples
  • 11 hours in the syllabus
0 of 7

Start here

Populations, samples and random samples

Lesson 1 of 7 · 12 min

Start the topic

Lessons

7
  1. Populations, samples and random samples (up next)

    Tell a population from a sample and a parameter from a statistic, and explain what makes a sample random and why randomness matters.

  2. Bias and procedures that ensure randomness

    Recognise the common sources of bias in samples and use random numbers, systematic sampling and stratified sampling to choose a sample fairly.

  3. Displays of real and simulated samples

    Recognise what random samples from uniform, Bernoulli, binomial and normal distributions look like, generate them with technology, and use their displays to estimate population features.

  4. The sample proportion as a random variable

    Calculate the sample proportion p^=X/n\hat{p} = X/n, see why it varies from sample to sample, and find its exact distribution for small samples from the binomial distribution.

  5. The mean and standard deviation of p̂

    Use E(p^)=pE(\hat{p}) = p and SD(p^)=p(1−p)/n\mathrm{SD}(\hat{p}) = \sqrt{p(1 - p)/n}, derive them from the binomial distribution, and work out how sample size and p control the spread.

  6. The normal approximation for p̂

    Recognise that p̂ is approximately normal for large samples, check when the approximation is reasonable, and use it with technology to calculate probabilities and quantiles for sample proportions.

  7. Simulation, standardising and when the approximation works

    Use repeated random sampling for different p and n to examine the distribution of p̂, standardise p̂ to a z-score, and explain why the normal approximation depends on both n and p.

  8. Topic quiz

    Ten questions across the topic, from foundation to complex unfamiliar. Your results feed your mastery.

About this topic

Nobody asks every Australian how they will vote, or tests every battery that comes off a production line. Instead you take a sample and use it to say something about the whole population. That only works if the sample is chosen well, so this topic starts with random samples: what makes a sample random, how bias creeps in, and the procedures that keep it out. You will also read graphs of real and simulated samples from uniform, Bernoulli, binomial and normal distributions.

Then the key idea of the topic: the sample proportion p^\hat{p} is itself a random variable. Take another sample and you get a different p^\hat{p}. You will find its exact distribution for small samples, learn that its mean is pp and its standard deviation is p(1−p)/n\sqrt{p(1-p)/n}, and see by simulation that for large samples its distribution is approximately normal. That approximation is what lets a pollster say "52 % ± 3 %" — the interval estimates of the next topic.

Before you start

Builds on Discrete random variables and Continuous random variables and the normal distribution.

  • The binomial distribution X∼Bin(n,p)X \sim \mathrm{Bin}(n, p), with E(X)=npE(X) = np and Var(X)=np(1−p)\mathrm{Var}(X) = np(1-p) (Unit 3).
  • Bernoulli random variables, with mean pp and variance p(1−p)p(1-p) (Unit 3).
  • The rules E(aX+b)=aE(X)+bE(aX + b) = aE(X) + b and Var(aX+b)=a2 Var(X)\mathrm{Var}(aX + b) = a^2\,\mathrm{Var}(X) (Unit 3).
  • Normal distributions, zz-scores and normal probabilities and quantiles with technology (Unit 4 Topic 3).
  • Dot plots, bar charts and histograms (Years 7–10).

What QCAA says you need to know7 dot points

Random sampling

3 hours

Sample proportions

8 hours

Subject matter from the Mathematical Methods General senior syllabus 2025 v1.3 (opens in a new tab), CC BY 4.0 © State of Queensland (QCAA).

The focus areas you can practise on their own. Each one maps to dot points above.

  • Populations and random samples

    Tell a population from a sample and a parameter from a statistic, and say what makes a sample random.

    Practise
  • Spotting bias in samples

    Name the source of bias in a sampling method (convenience, self-selection, undercoverage, non-response, question wording) and explain its effect.

    Practise
  • Procedures that ensure randomness

    Use random numbers, simple random sampling, systematic sampling with a random start and stratified sampling to choose a sample.

    Practise
  • Displays of random samples

    Read dot plots, bar charts and histograms of real and simulated samples from uniform, Bernoulli, binomial and normal distributions.

    Practise
  • The sample proportion

    Calculate $\hat{p} = X/n$, list its possible values and find exact probabilities for small samples from $X \sim \mathrm{Bin}(n, p)$.

    Practise
  • Mean and standard deviation of p̂

    Use $E(\hat{p}) = p$ and $\mathrm{SD}(\hat{p}) = \sqrt{p(1-p)/n}$, and explain how the spread depends on $n$ and $p$.

    Practise
  • Sample size and spread

    Find the sample size that gives a required standard deviation, and compare the spread of $\hat{p}$ for different sample sizes.

    Practise
  • Normal approximation for p̂

    Use $\hat{p} \approx N\big(p, \tfrac{p(1-p)}{n}\big)$ for large samples to calculate probabilities and quantiles with technology, and judge when it is reasonable.

    Practise
  • Simulation and standardising

    Use repeated sampling to examine the distribution of $\hat{p}$, standardise $\hat{p}$ to a $z$-score, and explain how the quality of the normal approximation depends on $n$ and $p$.

    Practise

Worked examples11

All 11

Common mistakes

Thinking a big sample fixes a biased method

A huge online poll of self-selected volunteers is still biased. Increasing nn shrinks the random variation in p^\hat{p}, but it does nothing about bias.

Confusing p and p̂

pp is a fixed (usually unknown) population parameter. p^\hat{p} is the proportion in one sample and changes from sample to sample. Only p^\hat{p} is a random variable.

Using the variance as the standard deviation

p(1−p)n\frac{p(1-p)}{n} is the variance of p^\hat{p}. Take the square root before standardising or entering σ\sigma into a normal probability calculation.

Using the standard deviation of X instead of p̂

np(1−p)\sqrt{np(1-p)} is the standard deviation of the count XX. Divide by nn to get p(1−p)/n\sqrt{p(1-p)/n} for the proportion p^\hat{p}.

Doubling the sample to halve the spread

The standard deviation depends on 1n\frac{1}{\sqrt{n}}, so you need four times the sample size to halve it.

Trusting the normal approximation when p is near 0 or 1

With n=50n = 50 and p=0.03p = 0.03, the distribution of p^\hat{p} is strongly skewed and the normal model even gives a positive probability to p^<0\hat{p} < 0. Check npnp and n(1−p)n(1-p) first.

Exam tips

  1. When asked to "identify a source of bias", name it and say which part of the population is over- or under-represented and which way the estimate is likely to be pushed.

  2. For "describe a procedure to select a random sample", give a method someone else could follow: the sampling frame, how each member is numbered, the random number command (e.g. RanInt#(1, 800)), and what to do with repeats.

  3. Write the distribution before you calculate: "p^≈N(0.3,0.022)\hat{p} \approx N(0.3, 0.02^2) because np=157.5np = 157.5 and n(1−p)=367.5n(1-p) = 367.5 are both large".

  4. Normal probabilities for p^\hat{p} are technology-active (Paper 2). Show the setup — the mean, the standard deviation and the bounds you entered — not just the answer.

  5. Keep the standard deviation unrounded on your calculator; round only the final probability, to the places the question asks for.

  6. If an answer says a proportion is negative or a probability is above 1, stop and check: it usually means the normal approximation is not appropriate or a variance was used as a standard deviation.

Where this shows up

  • Opinion polls: Newspoll and other pollsters survey around 1000–2000 voters and report a primary vote with a margin of error derived from p(1−p)/n\sqrt{p(1-p)/n}.

  • The ABS Labour Force Survey estimates the national unemployment rate each month from a sample of households, not a census.

  • Quality control: a factory tests a random sample from each batch and rejects the batch if the proportion of faulty items is unusually high.

  • Public health: Queensland Health surveys estimate the proportion of adults who smoke, are vaccinated or meet activity guidelines.

  • Ecology: the proportion of tagged fish in a recaptured sample estimates the size of a fish population in a dam.

Save your progress with a free account