A sample is a representative proportion that enables us to make informed statements about a population as a whole. We only look at samples because we cannot feasibly look at every single instance of something. Therefore all research involves sampling:

• Random sampling: In order to be able to make valid statements about the occurrence of something within a whole population, a sample must be representative, that is, typical of the population. Representative sampling or probability sampling is achieved with random selection. It is possible that some categorisation takes place before the sampling in order that the particular group investigated is better represented. This would then result in a stratified random sample.

• Purposive sampling: Sometimes a specific group may need to be compared with another. In this case sampling would be deliberate or purposive including just those members of the groups being compared.

• Stratified sampling: In this case the sample is selected from a more restricted group, and some categorisation takes place before the sampling in order that the particular group investigated is better represented. A highly stratified sampling technique would be one in which specific numbers for each category were selected, eg: 48 male informants (12 in each of the four social classes). Each group of 12 was further divided into a 10-12 age group, a 14-17 age group and a 30-55 age group.