Types of Data and Sampling · Cambridge syllabus 9.1 · Lesson 2 of 2
Sampling
Data collection, including sampling, is taught but not examined. A sample studies part of a population. Random sampling gives a fair selection mechanism; stratified sampling preserves relevant group proportions. Convenience or voluntary samples may be biased. Bigger samples do not fix a systematically biased selection method.
By the end, you can
- Understand the purpose of sampling (not examined).
- Use appropriate sampling methods (not examined).
- Recognise bias and limitations in samples (not examined).
- Discuss representativeness of a sample (not examined).
Sampling methods
This lesson is about collecting data, which Cambridge expects every course to teach but does not examine: there are no questions on sampling in the 0580 papers, and the practice questions here do not count towards your mastery score. A population is everyone or everything under study. A census surveys all of them; a sample surveys part. A census is accurate but usually too slow or expensive, so most work is done from samples.
A sample must be representative, or conclusions drawn from it do not apply to the population. Two things make it so: it should be large enough, and it should be chosen without bias.
In a random sample every member has an equal chance of selection, usually by numbering the population and using random numbers. In a stratified sample the population is divided into groups and each contributes in proportion to its size — surveying 30 from a year group of 200 out of a school of 1000 means taking 6 from that year.
Bias is what makes a sample misleading, and the standard examples are worth recognising: asking only people in one place, only volunteers, or only those willing to answer. Surveying shoppers on a weekday morning about employment will not represent people at work.
When you judge a sample, name the specific problem rather than saying it is "not fair", and suggest how to improve it. The same thinking helps when a question asks you to appreciate the limits of a conclusion drawn from data, which is examined.
Worth remembering
- Not examined, but part of the course: Cambridge expects data collection to be taught.
- A sample must be large enough and chosen without bias.
- Stratified sampling takes from each group in proportion to its size.
Worked example
For a school with 60% juniors, a proportional sample of 50 includes 30 juniors.
Practice
Warm-up · not marked
Check your understanding
Answer in your own words, then compare with the worked answer. Nothing here is saved or scored.
Worked answer
Club members are likely to exercise more, so the sample is not representative.
Check your method, units and explanation against it.
Marked practice
Set 1 of 4 · easiest first · Checked as you go, with the worked answer. Sign in or create a free account to save your progress.