What is Quota Sampling?
Quota sampling is a non-random selection of subjects from population subgroups that the researchers define. Researchers use quota sampling when random sampling isn’t feasible, and they want more control over who they select compared to other non-probability methods, such as convenience sampling.
Researchers define the subgroups necessary to achieve their study’s goals. Typically, the goals for quota sampling are either to represent the overall population or to facilitate comparisons between vital groups in the study. The investigators must define these groups so that participants meet the criteria for only one. These groups are called strata. During the project, researchers must learn enough about each subject to assign them to the correct stratum.
Strata are subgroups whose members are more like each other than the broader population. Researchers can create strata based on income, gender, and race, among many other possibilities.
The definition of quota is the number of participants for each group. Researchers determine these numbers during the planning phase before searching for subjects. When they select subjects, they do not use random sampling.
Stratified and quota sampling are similar because they divide the population into relatively homogenous subgroups. However, stratified sampling uses random selection, whereas quota sampling does not.
When to Use Quota Sampling
Probability sampling methods, such as simple random sampling, more effectively select representative samples, but they tend to be cumbersome, time-consuming, and expensive. Creating the sampling frame for the entire population is frequently a stumbling block by itself.
Quota sampling is an advantageous alternative when a study doesn’t have the time or resources to use random sampling. It’s a middle-ground approach. Quota sampling is not as effective at producing representative samples as random sampling, but it can be more effective than pure convenience sampling, which accepts all willing subjects. Learn more about Convenience Sampling.
Consider using quota sampling in cases where you could use convenience sampling, such as online and social media surveys, questionnaires at shopping malls and other crowded locations, acquaintances, and so on. However, instead of accepting all volunteers, you selectively include subgroups to meet your research goals better. If your subjects belong to a hard-to-find population, consider using Snowball Sampling instead.
Despite its aims, quota sampling is still a non-probability method; therefore, it has a greater potential for bias than probability methods. Researchers intentionally select participants, often based on convenience, which opens the door to bias. Consequently, the results are less generalizable from the sample to a population.
A quota sample is best for obtaining the big picture about opinions, concerns, situations, and ways of thinking—particularly when you need quick answers!
Quota Sampling Examples
Proportional quota sampling attempts to reproduce the characteristics of the entire population in the sample even though the researchers will not use random sampling. Researchers frequently use this method for surveys. To use this method, they must divide the population into relevant subgroups and then set quotas that retain the population’s group proportions.
For example, researchers might want to collect a quota sample that reflects the population proportions for ages, genders, and races. This process creates subgroups that use combinations of these three variables. For example, one subgroup would be female Latinas between 18 and 24. Repeat this process for all genders, races, and age groups.
In other cases, the researchers aren’t attempting to reproduce the population within the sample. Instead, quota sampling ensures they obtain enough participants from important subgroups to better understand the differences between them. This approach helps researchers compare the groups.
For example, if a researcher is studying the differences in attitudes between science fiction fandoms, they can set a quota for each franchise. This process ensures that the study has enough respondents for each franchise to compare them.
When using the non-proportional method, researchers have more freedom to choose the quotas that work for their study. However, these studies do not reflect the overall population.
How to Perform Quota Sampling
To perform quota sampling, do the following:
- Define the strata that are important to your study.
- Determine the quota for each stratum.
- Recruit until you fill each group’s quota.
Step 1: Defining the Strata
When defining the strata, use your subject-area knowledge to determine which ones are important to your study. Generally, you include strata in your quota sample to either ensure representation or compare groups. Your quota sampling goal affects both the strata and quotas.
You’ll need to define the subgroups in a mutually exclusive manner. Each subject can qualify for only one.
Step 2: Determining the Quotas
When determining each group’s quota, keep your research goals in mind. First, choose the total sample size across all strata. This number depends on the time and resources available. Then determine the quota for each subgroup.
If you want the quota sample to reflect a population, you’ll need to know each group’s population proportion, and then choose group numbers that match those proportions. That process involves research!
For example, if you’re surveying students at a university and want the proportion of majors in the sample to reflect the university’s population, you’ll need to assess administrative records. If statistics majors comprise 10% of the population, and you want a total sample size of 1,000 participants, then you’ll need to use a quota of 0.10 X 1,000 = 100 for statistics majors. Repeat for the other majors.
On the other hand, if your quota sampling goal is to facilitate comparisons between groups, you’re freer to choose numbers that suit you. Specifying equal quotas for all comparison groups is often a good starting point. For example, if you want to compare attitudes between fans of differing science fiction franchises, you might include equal numbers for each fandom:
- Doctor Who: 300
- Star Trek: 300
- Star Wars: 300
Step 3: Recruiting
Finally, recruit and fill up those quotas! In addition to asking your primary questions of interest, you’ll need to gather enough information to assign each subject to the appropriate strata. Recruiting for quota sampling often entails the same methods as convenience sampling, except you don’t use all volunteers in the study. As the quotas fill up, you’ll start turning people away.
For the campus survey, place recruiters at busy campus locations.
To conduct the science fiction fandom survey, recruiters might attend science fiction conventions dedicated to each franchise.
Advantages and Disadvantages of Quota Sampling
Compared to probability methods, quota sampling has the following advantages and disadvantages.
Quota sampling tends to be cheaper, quicker, and easier than probability methods. You won’t need to create the entire population list and be forced to find the relatively dispersed subjects that random chance picks. Instead, you can find participants in convenient locations using expedient methods. However, unlike convenience sampling, this method gives you control over the sample’s constituents.
If you’re comparing specific groups, they are built right into the quota sampling process. You know you’ll have sufficient numbers.
If you have a small budget or limited time but need to control the representation in the sample, consider quota sampling.
Unfortunately, the fact that quota sampling does not use random selection increases the potential for bias. Ultimately, the researchers pick who is in the study, and convenience is generally a factor. The higher potential for sampling bias reduces your ability to generalize from the sample to the population. Learn more about Sampling Bias.
Additionally, while you might stratify your quota sample on crucial characteristics to increase representativeness in those areas, it might not be representative for strata you don’t include. For example, if you base your strata on gender, race, and age, the sample might not accurately represent income or religion.