Multi-stage or cluster sample
As the name implies, this involves choosing a final sample by stages. It might be that a number of counties are randomly chosen; two each from England, Scotland, Wales and Northern Ireland, and then two towns are randomly selected from the counties and ten streets from each town. Finally householders are randomly chosen from each street to answer the survey questions. Such a sampling method is quite complex and, though it might produce a representative picture of British households, it is far more trouble to organise than surveying people in one town.
In contrast to multistage sampling, this is a straightforward method often adopted for market research. The interviewer has been instructed to question, for example, a hundred women of childbearing age. Women who look the right age are stopped in the street and asked to answer questions, while those who look too old or young are ignored. A problem is that the sampling frame is less representative than it appears to be. Particular groups, such as those with dependents or people busy working, are less likely to be out in the street.
Sociologists who are lecturers or teachers may persuade their captive audience of students to answer questionnaires on the spot. While resulting in 100% response rate, the sample made up of sociology students is unrepresentative of the population as a whole. Amateurs often ask only their friends and relatives to answer questions, usually people drawn from a very narrow social spectrum.
Sociologists are not always in search of a representative sample. Sometimes they seek exceptional cases to undermine a generalisation. For example in response to G. P Murdock’s contention that the nuclear family was an integral part of living arrangements all over the world, E. Kathleen Gough produced a case study of the Nayar found in Kerala, India in the eighteenth century, who lived in an entirely different way. A case study is a detailed investigation of a particular group or institution, whether it is typical or not, usually using a variety of research methods. In this case Gough had to use secondary sources for her historical study.
This term refers to a detailed study of a particular group. (Despite the name, it is not necessarily a study of ethnic minorities. The word literally means writing about a group of people or culture). As with a case study, the group or institution does not necessarily have to be representative. It might be studied because it is exceptional and therefore particularly interesting, such as A.S. Neill’s progressive school, Summerhill. At other times, the researcher might imply that the group studied is typical and generalise from it, as Paul Willis did of ‘the lads’ in Learning to Labour: How Working Class Kids get Working Class Jobs (1977). An ethnographic study usually involves fieldwork, direct intensive observation on site, often through participation. Other methods might also be employed, such as interviews and consulting any available secondary sources such as school records.