What kind sampling based on probability?

Answer 1

Probability sampling, according to which, a member of the population has the same probability of being included in the sample as any other member. Equivalently, each sample of a given size has the same probability of being chosen.

Stratified and cluster sampling are variations on this idea. In stratified sampling, the population can be divided up into strata such that members within each stratum are more like each other than across strata. One example may be school pupils in different year groups. A sampling scheme could assign a number to be sampled from each stratum (perhaps according to how large that group is) and then, within that stratum, to use simple probability sampling.

Cluster sampling is used when the entire population can be split up into clusters. Clusters are selected using probability sampling. Then a census is used within each cluster. For example, if you wanted to sample schools across the country, a simple probability sample would result in schools all over the country and the travelling costs (and time) would be prohibitive. Instead, you divide the country up into regions and take a probability sample of these regions. You end up visiting every school within the few selected regions.