Systematic sampling is a way of conducting research that determines how to select members of a population that will be studied. Many research efforts concentrate on getting a random sample, where every member of the population under study has the same chance of being chosen. Another choice is taking a simple random sample, where every group of the predetermined size has the same chance of being selected. One alternative is systematic sampling, where researchers select a starting member of the group, and then use this as a means of selecting all other samples.
Anyone who has ever participated in team sports in a physical education class has probably experience systematic sampling. A coach heads down the line of students, counting every other student, and breaks up the line into two teams. Essentially the coach begins with a starting student, and takes every other student with him, creating a sample of the original group that will become team one or team two.
It would usually be impractical to pick “every other” of a large group. Instead, researchers determine the kth element. K is defined as the number of elements in the population that will be skipped in between picks, and it must remain constant for the whole sample selection process. If someone taking a sample decides that the sample begins with one and every 50th element thereafter will be drawn as part of the sample, k is 50. The sampler will pull or test 1,51, 101, 151, and so on, until reaching the end of the group.
In statistics, a simple random sample is often preferred because there are many operations that require one, so that more information about something can be assessed. Systematic sampling is highly useful and well-suited to the purpose of determining information about a population for reasons such as quality control. It is important to point out that the sample isn’t fully randomized, though it may be one of the better approximations of randomness.
Much of the accuracy of a systematic sample depends on choosing a sample size representative of the population. This means k has to be small enough to create an adequate sample size that tells researchers or testers about what is happening in the general population. K cannot be too small or the group it selects is very large and may be impractical and expensive to test. It should be noted that a systematic sample can be of people, animals or things, depending upon what is being tested.
Systematic sampling appears frequently in manufacturing. Many factories automatically inspect or draw from the line every kth product. Some food manufacturers, in particular, use this as a selling point to demonstrate high standards of quality control.
