Convenience Sample

Nonprobability sampling is used by researchers to find members of a population pool when this population cannot be enumerated to allow for a probability sample to be constructed. There are four general categories of nonprobability sampling: convenience, quota, purposive, and snowball sampling.

A convenience sample is a selected group from a particular population that is chosen based on their accessibility to the researcher. Examples of convenience samples include selecting neighborhood residents by inviting participants from those entering a supermarket, selecting grade school students from the local elementary school, and selecting undergraduate students from a freshman-level class.

A quota sample is a convenience sample in which the researcher seeks to control some variability in his or her sample. As in a probability sampling—stratified sampling method—a researcher may preselect the proportions of certain strata that he or she desires in the final sample. For example, a researcher who desires that 50% of the final sample of 100 students be male may recruit subjects from a freshmanlevel class of undergraduates with a different proportion of male students but restrict participation among males once 50 agree to participate, focusing solely on the inclusion of females until 50 are included in the sample.

A purposive sample is a convenience sample whose population parameters are defined by the purpose of the research. For example, a researcher may want to understand more about undergraduate students who work full-time. To obtain potential participants, the researcher may place fliers throughout the school asking specifically for undergraduates who are currently holding a full-time job.

A snowball sample is a convenience sample in which the researcher identifies a smaller pool of accessible members of the particular population and requests that those members identify other appropriate potential members. This method is similar to a purposive sample selection but is useful if access to the population pool is particularly challenging for the researcher but less so for members of that population. For example, a researcher interested in organ donation may use a known organ donor who is associated with donor support groups to facilitate access to the members of these groups.

Although it is impossible to determine the probability of selection in a nonprobability sample, and thus a researcher cannot statistically measure the representation of the selected population, this does not necessarily mean that the sample is not representative. Whereas nonprobability samples cannot be assessed quantitatively, they can and should be assessed qualitatively with regard to the true sampling frame of the participants in the study. Often, a researcher can use the demographic characteristics of his or her sample to explore how well they compare with the population of interest. Additionally, a researcher should be thoughtful as to why the particular group was accessible and if this in any way may influence the findings of the study.