Quota Sampling

Sampling designs are typically separated into two fundamental types: probability and nonprobability. In probability-based sampling, each member of the population has a known, nonzero chance of being selected. Nonprobability samples are generated from more subjective criteria of the researcher, such as personal experience, convenience, and volunteers. Quota sampling is one form of a [Page 1357]nonprobability or judgmental sampling design used to acquire data from population subgroups.

In quota sampling, participants or locations are selected nonrandomly according to a fixed quota or percentage of the population based on one or more characteristics. The quota selected may be proportional or nonproportional to the actual population distribution. In order to obtain a representative sample, a proportional quota sampling design requires a priori knowledge about the underlying population characteristics. Nonproportional quota sampling is less restrictive, as the researcher specifies the percentage of data elements to be sampled from each subgroup independent of the population.

The quota sampling procedure is outlined as follows: (a) divide the population into subgroups that are exhaustive and mutually exclusive (i.e., all data points occur in one and only one division), (b) stratify the population into classes based on one or more characteristics (e.g., gender) and determine the proportion in each class, (c) pick a sample size, (d) select a quota for each subgroup that may either be proportional or be nonproportional to the population, and (e) collect data points until the quotas are completed.

For example, a psychologist is interested in the attitudes of people toward therapy based on U.S. political party affiliation in their area. From voter registration data, the population is 60% Democrat, 35% Republican, and 5% Independent. A sample size of 200 adults is chosen with political affiliation as the criterion. For a proportional quota sample, the psychologist selects 200 individuals who represent the political party distribution of their area (i.e., 120 Democrats, 70 Republicans, and 10 Independents). If the psychologist is more interested in the attitudes of Republicans, the quotas could be altered to a nonproportional sample with a higher sample size for Republicans than is found in the general population (i.e., greater than 35%); however, the sample would be less representative of the population.

Quota sampling offers several advantages over more complex probability sampling methods. For primary data collection, quota sampling is relatively inexpensive, quick, and simple. Researchers can also ensure data are collected from all subgroups for a given set of characteristics, thus guaranteeing smaller groups are represented in the sample. Yet, a major disadvantage of quota sampling is that the selection process is nonrandom and subjective, especially for nonproportional samples. Consequently, difficulties arise in determining the sample error or making inferences about the general population from the sample.

The shortcomings of quota sampling are often reduced in stratified random sampling, the probability-based alternative whereby an element of randomness is introduced. Using stratified random sampling, each data member of the population has the same probability of being selected as any other member throughout the sampling process. In contrast, quota sampling reduces the chance of a data member being chosen as the quota is filled.

...