Simple Random Sampling

Simple random sampling is considered a good standard sampling approach, however it requires a list of all units in the population. This is achievable for groups such as student cohorts or club memberships, but impossible for groups with unclear numbers such as abuse victims or cancer sufferers. In simple random random sampling, each unit must have equal chance of selection.

Stratified Sampling

In stratified sampling, the population is divided into strata (groups), based on a specific characteristic. This is typically done to ensure all important groups are sampled, including minorities which random sampling may exclude. It requires information about the prevalence of each group and random samples should be taken from each sample proportionate to its' size.

Systemic sampling

Each $yth$ member of the population is sampled. The starting point is typically chosen at random.

Cluster Sampling

The population is divided into clusters, or groups. Then entire clusters are randomly selected and all members of the cluster are included in the sample.

Multistage Sampling

In the first stage the population is divided into clusters and a sample of these clusters is selected, usually randomly. In the second stage a sample from these clusters is selected, usually randomly.

back to contents | next lecture