This module covers how to define a population and sampling frame, and when to use probability designs (simple random, systematic, stratified, cluster and multistage) or non-probability designs (purposive, quota, convenience, snowball and theoretical). Sampling decisions set the limits of what you can generalise, so examiners scrutinise both the method and the sample size justification.
Free previewLesson 1 · Populations, sampling frames and units of analysisPreview video to be added
Watch the first lesson of every module free. The remaining lessons, exercises and live sessions open when you enrol.
What you will be able to do
Define a target population, accessible population and sampling frame for your study
Select and justify a probability or non-probability sampling design
Justify sample size using an a priori power analysis or, for qualitative work, information power and saturation arguments
Explain sampling bias, non-response bias and the design effect of cluster sampling
Key terms
Sampling frame
Stratified sampling
Cluster sampling
Purposive sampling
Snowball sampling
Statistical power
Design effect
Data saturation
Hands-on practice
Draft a sampling plan for your study that names the frame, the design, the inclusion and exclusion criteria and a sample size justification (for example from G*Power).
Common pitfalls this module fixes
Describing a convenience sample as 'random'; name the actual method and discuss its limits on generalisation.
Picking a sample size by rule of thumb; base it on an a priori power analysis with a stated effect size, alpha and power, or on a reasoned qualitative argument.
Ignoring clustering when sampling schools, wards or firms; account for the design effect in sample size and use multilevel or cluster-robust analysis.