Cluster Sampling

Question 1

What is characteristic of clusters in cluster sampling?

Answer 1

a) they are representative of the wider population
b) should be heterogenous

Question 2

What is the cluster sampling procedure?

Answer 2

1) Identify suitable clusters;
2) Take a random sample of clusters
3)Take all population elements in each selected cluster into the sample (i.e. one-stage cluster sampling - the simplest case);
4) Or, take a further random sample from each selected cluster (i.e. two-stage cluster sampling – conducted if clusters are too big to examine in full).

Question 3

How can cluster sampling be extended?

Answer 3

a) multi-stage cluster sampling = more than one level of random sampling
b) cluster sampling is often combined with stratified sampling, stratified by geographical location

Question 4

What are the advantages of cluster sampling?

Answer 4

a) convenient and low cost
b) avoids need for sampling frame of all population elements
c) most effective when within cluster sampling variabilities are high relative to population variance