Sampling techniques

In one line: You can't study everyone, so you pick a sample — and how you pick it shapes who your findings apply to.

A sample stands in for the whole population. Sampling is how you choose that group, and it decides how far the results can be trusted to generalise.

This links to the concept of bias: a poorly chosen sample can be unrepresentative, so the findings are biased and don't apply to the wider population (poor generalisability).

Memory hook: A sample stands in for the population. Choose it badly and your results are biased.

Key idea: Techniques trade off between how representative and how practical they are.

Take studying student stress. Grabbing friends in the library is opportunity (fast, biased); putting up a sign-up sheet is self-selected; drawing names from the whole roll is random (most representative). The technique affects how much bias is in the sample.

Go further — higher-level insight: No sample is perfect — name the bias. Opportunity samples over-represent whoever's around; volunteer samples attract the keen and confident. A strong answer doesn't just name the technique — it says who gets left out and how that limits generalisability.

Key idea: The best technique depends on the trade-off between representativeness and practicality.

So representative techniques (random, stratified) generalise better but take more effort, while practical ones (opportunity, self-selected) are fast but risk bias. A good evaluation names the likely bias of the chosen technique and how it limits generalisability.

Watch out: A biased sample limits generalisability. If your sample over-represents one type of person, your findings may only apply to them — not the whole population.