1. A small quantity of a commodity, etc., selected to represent the bulk of a quantity of goods. See sale by sample.
2. A small quantity of a product, given to potential buyers to enable them to test its suitability for their purposes.
3. A group of items or individuals selected from a population to represent the characteristics of the population as a whole. Samples are often used in marketing research because it is not feasible to interview every member of a particular market; however, conclusions about a market drawn from a sample always contain a sampling error and must be used with caution. The larger the sample, in general, the more accurate will be the conclusions drawn from it. In quota sampling the composition of the sample reflects the known structure of the market. Thus, if it is known that 60% of purchasers of household DIY products are men, any sample would reflect this. An alternative sampling procedure is random or probability sampling, which ensures that everyone in a particular market has an equal probability of being selected. Although more rigorous than quota sampling, it is also more expensive and difficult to implement. See also cluster sampling; judgmental sampling; stratified sampling; systematic sampling.
Sampling is also important in auditing, as it is not usually feasible to inspect all the available documentation.