Читать книгу Sociology - Anthony Giddens - Страница 89

Often sociologists are interested in the characteristics of large numbers of individuals – for example, the political attitudes of the population of Australia. It would be impossible to involve all 25 million people directly, so in such situations researchers engage in sampling – concentrating on a small proportion – a sample – of this overall population. One can usually be confident that the results from a population sample, as long as it is properly chosen, can be generalized to the total population. Studies of only 2,000 to 3,000 voters, for instance, can give a very accurate indication of the attitudes and voting intentions of the entire population. But, to achieve such accuracy, a sample must be representative – that is, the group of individuals studied must be typical of the population as a whole. Representative sampling is more complex than it may appear, and statisticians have developed rules for working out the correct size and nature of samples.

A particularly important procedure used to ensure that a sample is representative is random sampling, in which a sample is chosen so that every member of the population has the same probability of being included. The most sophisticated way of obtaining a random sample is to give each member of the population a number and then use a computer to generate a random list from which the sample is derived – for instance, by picking every tenth number.

In fieldwork, researchers have to get close to the communities they are studying, but not so close that they lose their relatively detached sociological eye.

There are other types of sampling used by sociologists. In some types of research, it may be necessary to use convenience sampling. This means taking your sample from wherever you can! Because convenience sampling is less systematic and rigorous than other types, the results it generates have to be treated with caution. Nonetheless, in applied research or studies of hard-to-reach social groups who may be reluctant to come forward – for example, substance users or people who self-harm – it may be the only practical way of gathering an adequate sample. Without convenience sampling, the voices of some social groups may just not get heard. Similarly, snowball sampling, in which existing participants are used to recruit other participants via their own network of contacts and friends, is a tried and tested method of gaining access to a larger sample than would otherwise be possible.