MathsStatistics › Sampling and the large data set

Sampling and the large data set

You can rarely measure everyone, so you measure some and argue about the rest. This lesson is about when that argument holds up: what each sampling method buys you, what it costs, and what the exam's own data set looks like.

Year 12-13EDEXCEL 9MA0 S1

IN THIS TOPIC

  • Define population, census, sample and sampling frame, and say when a census is impractical.
  • Describe simple random, systematic, stratified, quota and opportunity sampling, with one advantage and one limitation of each.
  • Know the shape of the large data set: the weather variables, the stations and the periods.

WHAT YOU PROBABLY THINK

A larger sample automatically makes the results representative.

Why sample at all

A population is every member of the group you care about; a census measures all of them. Sometimes that is fine: a teacher can ask every student in one class. But a census of a large population is slow and expensive, and sometimes impossible on principle, because the measuring destroys the thing measured. You cannot crash-test every car.

So you take a sample: a subset, measured fully, standing in for the rest. The whole subject of statistics is the argument from sample to population, and the argument is only as good as the way the sample was chosen. The list you choose from, every population member with an identifier, is the sampling frame.

The random methods

A simple random sample of size n gives every possible group of n an equal chance: number the frame, then use a random number generator or lottery draw. It is the gold standard for fairness, and the standard against which everything else is judged. Its cost is practical: you need a full frame, and the chosen members may be scattered and hard to reach.

Systematic sampling takes every kth member of the frame after a random start: for 50 from 1000, k = 20, start anywhere in the first 20. Quick to run on a long list, and still random, provided the list has no repeating pattern that lines up with k.

Stratified sampling: a school of 240 people split into strata of 120, 80 and 40, and a sample of 30 taking 15, 10 and 5 from themYear 12: 120take 15Year 13: 80take 10Staff: 40take 5same fraction of every stratum: 30 out of 240 is one in eight
FIG. 1Stratified sampling in one picture: each stratum sends the same fraction, so the sample is a scale model of the population.

Stratified sampling splits the population into groups that matter, strata, and takes a simple random sample from each, sized in proportion. The number from each stratum is

stratum sizepopulation size × sample size

so a school of 240 sampled 30 strong takes one person in eight from every stratum. The sample inherits the population's structure by construction; the price is needing to know the strata and their sizes in advance.

The non-random methods

Quota sampling keeps the proportions but drops the randomness: an interviewer fills quotas, so many from each group, choosing whoever comes to hand until each quota is full. No frame needed, fast, cheap, and the standard tool of street surveys. But the interviewer's choices can smuggle in bias that no arithmetic can remove.

Opportunity sampling (or convenience sampling) simply takes whoever is available: the first twenty people through the door. It is the quickest method and the weakest, because availability is rarely unrelated to what you are measuring. Ask about exercise habits at a gym entrance and the sample answers a different question.

WORKED EXAMPLE

Choosing and using a method

A sixth form has 600 Year 12 and 400 Year 13 students. Explain how to take a stratified sample of 40 by year group.

The fractions come first: Year 12 supplies 40 × 600/1000 = 24 students and Year 13 supplies 40 × 400/1000 = 16.

Then each group of names is numbered and a simple random sample of the right size is drawn from each, using random numbers. Both halves of the answer earn marks: the proportional arithmetic, and the random selection within each stratum.

The large data set

Edexcel's large data set is real weather-station data, and the paper assumes you have worked with it. It records daily values, temperature, rainfall, sunshine, wind speed and direction, cloud cover, visibility and pressure, for the months May to October in two years, 1987 and 2015.

The Edexcel large data set: five UK weather stations and three overseas, recorded May to October in 1987 and 2015UK stationsCamborneHeathrowHurnLeemingLeucharsoverseasBeijingJacksonvillePerthMay to October1987 and 2015daily temperature, rain, sun and wind
FIG. 2Eight stations, two summers-to-autumns, twenty-eight years apart: the raw material for every large data set question.

The stations matter: five in the UK, Camborne, Heathrow, Hurn, Leeming and Leuchars, and three overseas, Beijing, Jacksonville and Perth. Perth is in the southern hemisphere, so its May-to-October window is winter, a favourite exam trap. Rainfall entries of tr mean a trace, less than 0.05 mm, and some entries are missing: real data needs cleaning before it needs summarising.

THE EXAM BIT

  • Name the method a description matches, then give an advantage and a limitation in context, not from a memorised list: the marks are for this sample, this population.
  • Stratified calculations are two marks of arithmetic: stratum over population, times sample size.
  • For the large data set, know the stations, the two years, the May-to-October window, and that Perth's seasons run opposite to the UK's.
  • “Random” has a technical meaning: every member (or every group of n) equally likely. Quota and opportunity sampling are not random, whatever their proportions.

CHECK YOURSELF

A factory's 1200 workers are 900 full-time and 300 part-time. Describe how to take a sample of 60, stratified by contract type.

Show a hint

Fractions of the population first, then random selection inside each stratum.

Show the answer

Full-time supplies 60 × 900/1200 = 45 workers; part-time supplies 60 × 300/1200 = 15.

Number each group's members and draw a simple random sample of the required size from each, for example with a random number generator.

A sample speaks for its population only if chance, not convenience, chose it.

Stratify when you know the structure: the same fraction from every stratum keeps the sample a scale model.

WORKBOOK

Printable practice for this topic: original exam-style questions with room to work, and a fully worked answer book. Free to use; please do not redistribute or sell.

CHECK YOUR PROGRESS

Rate how confident you feel with each objective for this lesson. Ratings are saved in this browser, on this device only.

  • Define population, census, sample and sampling frame, and say when a census is impractical.
  • Describe simple random, systematic, stratified, quota and opportunity sampling, with one advantage and one limitation of each.
  • Know the shape of the large data set: the weather variables, the stations and the periods.

No animated video for this topic yet; these notes stand alone.