Ohne Titel

Seite 1

How do sociologists collect a sample? There a number of sampling techniques used by sociologists in order to obtain a sample. The sampling techniques fall into two categories –

1. Random sampling: simple; this is the most straightforward way of selecting a random sample. It is the same as drawing names out of a hat! This, however, is only practical for small populations. For larger populations, researchers use computers to select their random sample. The potential problem with simple random sampling, however, is that, by chance, the random sample may not be representative of the population stratified; imagine that a sociologist is interested in researching the attitudes of members of a group involved in environmental issues. She does not have sufficient time or money to interview for population, so she will have to interview a sample of members. She has a list of members, and has advanced knowledge of their age and gender, and wants the sample to reflect the age and gender characteristics of the population. In order to achieve this she will divide the population into sub-populations, in this case according to age and gender. A sample is then drawn randomly from each sub population in proportion to numbers in the population as a whole. It's a bit like sorting the names into a number of hats and then drawing them out randomly. cluster; if the survey population is spread out across the country, it would prove relatively expensive and time-consuming to interview a randomly selected sample. Instead of going all over the country to get information from the sample, the researcher could use cluster sampling. This involves selecting certain areas at random, for example London and Manchester, and then selected a sample of people from these areas. The sample will therefore be located in clusters around a geographical area so it is easier to collect the data. 2. Non-random sampling: systematic sampling; systematic sampling involves taking every nth name from the sampling frame, e.g. systematically taking every 10th name from the electoral register. Imagine that the population consists of 1000 people and a sample size of 100 is required. Using systematic sampling, the researcher would select a number at random between 1 and 10. If this turned out to be 5, then the 5th, 15th, 25th, name -- and so on up to the 995th name -- will be selected from the sampling frame. This would give the required sample size. snowball sampling; a sociologist may be interested in studying a population for which is no sampling frame, for example people who fraudulently claim DSS benefits while working in paid employment. Without a sampling frame the researcher would not be able to select a random or systematic sample. Using the snowball sampling technique, the researcher would begin by making contact with one member of the population, gradually gaining his or her confidence until that person is willing to divulge the names of others who might co-operate. In such a way, the researcher would obtain a sample, although it is unlikely to be a representative one. quota sampling; this is a technique favoured by market research companies which employ and train paid workers to interview people on the street. Each interviewer is told to interview an exact number (or quota) off people from categories all groups such as females, pensioners or teenagers, in proportion to their numbers in the population as a whole. This method is not random and it depends on the interviewer's ability to spot the right type of person to fill their quota. One problem with quota sampling is that the interviewer may, in practice, make a mistake. Another problem is that the interviewer may, at the end of a long hard day, fiddle the quota. Quota sampling will only work well when the researcher knows a lot about the population of study.