Home » Eng Finance » S » Sample

Sample

In everyday terms a sample is a quantity of something which has been selected as representative. A wine taster takes a sample of a particular type of wine and tastes it in order to evaluate the whole vintage. In economic and social research, however, a sample is given a somewhat more restricted meaning: rather than being just any ‘part’ of the ‘population’, it must be selected by certain statistical methods designed to ensure that it is really representative of the population as a whole. Thus, the term ‘sample’ should strictly only be applied to a properly chosen set of items. (For discussion of what constitutes ‘proper’ choice, see quota sample; random sample; simple random sample; stratified sample.) The purpose of a sample is to provide information about the population from which it is selected. Since the sample is only a part of the population, it follows that we would expect some inaccuracy. For example, by questioning a sample of voters, it might be found that 30 per cent of the sample intended to vote for Presidential candidate A, 50 per cent of the sample intended voting for Presidential candidate B, and 20 per cent for other candidates. On the other hand, if an election were held, we might well get shares of 33, 49 and 18 per cent respectively, simply because a sample is unlikely to give a perfect representation of the population. However, if the sample is properly designed, the size of the likely error is kept to a minimum and it is possible to calculate and specify its size. Furthermore, it is possible to find how varying the number of items included in the sample affects the size of the likely error. The advantages of taking a sample, even if the information gathered is not likely to be perfectly accurate for the whole population, are often substantial. In the extreme case, where evaluating the sample destroys the items (tasting wine, testing a machine or component to destruction), it is quite clear that the relevant information can only be gained from a sample. In economic and social research, the relevant consideration is the relation between the size of the sample and the cost and quality of the information-gathering procedure. The larger the sample, the greater is the number of questionnaires, interviewers, etc., and the greater the cost of analysing the data.

Reference: The Penguin Dictionary of Economics, 3rd edt.