Skip to content

sample

(NL: steekproef)

A sample is a subset from a population on which the research is conducted.

Under certain circumstances, you may assume that results obtained in the sample also apply to the population as a whole.

A "good" sample has several properties:

  • the sample is random
  • the sample is sufficiently large
  • the sample is representative for the population

Ideally, a researcher would do the following to draw a sample from a population:

  • first and foremost, the population must be well defined
  • then you can compile a sampling frame, i.e. a list of all individuals in a population
  • then the researcher randomly selects a number of individuals

It is often difficult or impossible to set up a sampling frame. In such cases, the researcher chooses a different sampling method and tries to approximate the properties listed above as closely as possible within the available time and budget.

Even under the best of circumstances, the result within a sample is likely to deviate from the "true" value in the population, although we hope we can approach it sufficiently. So by taking a sample we introduce errors.