Sampling Is Best Utilized When

¶ … sampling is best utilized when a researcher knows which subsection of a given population he or she wishes to analyze and is limited on time or budget, or when it is not necessary to have a very accurate sample.

Quota purposive sampling uses control dimensions to determine which persons are surveyed, and these dimensions reflect the information the researcher wishes to obtain. For example, a study is being done on a university campus to determine freshmen and sophomore student lunchtime fast food preferences. The researcher knows that student body is 48% male and 52% female, and that the population size of both the freshman and sophomore classes is exactly the same. A researcher might select control dimensions of gender and undergraduate class. The researcher will then ask the canvassers to only survey freshman and sophomore undergraduates. The survey is complete once the surveyor has collected the adequate number of samples for each of the control dimensions. If the sample size is 100, the canvasser will collect 24 male freshman samples, 26 female freshman samples, 24 male sophomore samples, and 26 female sophomore samples.

However, this example is rather simplistic since there are only two controls and, for illustrative purposes, the undergraduate class sizes are made exactly the same. For the above example we are able to use precision control to get precisely the number of representative samples that are required (Cooper & Schindler, 2008). In another instance of quota sampling, a researcher may require four controls: age, gender, undergraduate class, and political affiliation. The researcher may know that the population under study is 25% 18 years old, 52% male, 30% freshman, and 60% liberal. The researcher would then ask the surveyor to interview students to reflect this demographic. In frequency control, the type of quota purposive sample being employed in this example, the surveyor would attempt to get a sample that comes close to the known percentages in the population (Cooper & Schindler, 2008). With frequency control, an exact sample is not required, only that the percentages surveyed are close to the known percentages in the population.