Quota sampling: Definition, types, and examples

Sampling is a crucial process that can make the difference between credible and unreliable research. 

In general, sampling is the process of selecting a representative subset of people from a larger population for research purposes. Matthew Ovington, cofounder of Research Connections, a consumer research agency, uses an example to illustrate: “If you are conducting research concerning all females in the U.K., you don’t need to speak to the millions of females that represent the entire population. What you need is to find a subset of people that replicates the larger group.” 

This means making sure there’s a spread of all the different factors that make up the audience. In this example, females in the U.K. span age groups, ethnicities, regions, and locations as well as different behaviors and attitudes within the research topic. The key is to get a spread that exactly replicates the total population. 

But that presents a challenge: If the sampling profile is incorrect, you may end up with bias. So, the goal of sampling is to ensure you obtain a perfect replica of the research population without introducing any bias. 

While there are several sampling methods, quota sampling is particularly helpful for accurately replicating a larger population because it takes into account all the proportions of the population.

The background on quota sampling

Quota sampling is a type of non-probability sampling that involves selecting people or things within a population on the basis of their characteristics. 

In our example, we may use a nationwide census to identify the age and regional differences among females in the U.K. We may find that 20 percent of them are based in Scotland, 60 percent in England, 10 percent in Wales, and 10 percent in Northern Ireland. When researchers are armed with this information, their goal is then to select a sample of females with the same quota structure.  

The way quota sampling works

You could think of quota sampling as pushing people through a funnel, explains Ovington. 

“At the very top is the total population from which you can pool samples. Only research subjects who match the profile you’re building will pass through the funnel. If you’re using a questionnaire to collect responses for the U.K. study example, the first question would be to determine the respondent’s gender. Any respondent who is male is immediately screened out because they would not fit the profile,” he says.

“The next question could be about age, and you may code up their age brackets based on the spread from the national census. The goal as you collect responses is to match the quotas you’ve set for that research to the people coming through,” adds Ovington.

If we’re keeping track of responses in a two-column spreadsheet, one column would be static — the quotas set based on information from the census — while the other column would be calculated live as data is collected. The goal is to have a similar quota breakdown in the live column as in the static column at the end of the sampling process. 

For example, say your quota for females from Scotland is 20 percent. As you collect responses, you’ll need to calculate the percentage of respondents who are from Scotland. Once you reach the set quota, you’ll stop accepting further responses from females in Scotland. 

The same process is applied for other factors in the study. If you don’t limit the responses from females in Scotland, then they’d make up a larger proportion of your sample size than they do in the total population, which means your sample wouldn’t be an accurate representation of females in the U.K. 

Controlled and uncontrolled quota sampling

Quota sampling may be controlled or uncontrolled. Controlled sampling involves setting definite quotas based on the breakdown of the population. Here, quotas are usually inflexible and must be met as defined. They work best for quotas set against demographic characteristics like age and gender, such as the females in the U.K. study example. 

Uncontrolled quota sampling involves setting flexible quotas for monitoring purposes and adjusting the quotas as the data builds up. 

For example, if you’re conducting research on dog owners who feed their dogs premium food, you may have some idea about the characteristics of the target demographic. But it would be difficult to know the exact audience breakdown in terms of gender, age, or region. For that type of research, it may be more useful to set a “soft” quota that you can then adjust as you collect data.

Steps in quota sampling

To conduct quota sampling effectively, Ovington says you would typically follow the steps identified below:

Step 1: Find research that already exists for your population

To set quotas, you need credible research that gives a good breakdown of your target population. For example, this could be any kind of census report or information from a government database. It’s important to make a judgment call on whether the data is credible or not. For commercial research, you could rely on existing research conducted by a business or organization. 

Step 2: Set your quotas

Once you have credible research on which to base your quota, the next step is to use it to set quotas. 

It’s best to set quotas around facts, such as demographics like age and gender.  Avoid setting quotas based on attitudes or behaviors because those are likely to change. For instance, setting a quota against the frequency with which people drink alcohol in a week isn’t likely to yield great results for a couple of reasons. 

First, people are generally poor at recalling things and may provide wrong answers. Secondly, people may be ashamed of their drinking habits and quote lower numbers to avoid judgment. If your quota is set against these kinds of characteristics, it will be difficult to conduct tracking studies because you’ll get different results every time. 

If a third party will conduct the research, you’ll need to list out the quotas in a quota spec for whoever will handle the sampling.

Step 3: Conduct the survey

This is where the real work happens. Once you’ve set your quotas, the next step is to recruit people for the survey. Say you’re working with a sample size of 1,000 people and the following quotas are set for age:

You may find that the quota for the older age bracket fills up quickly because older people are generally more responsive to surveys than younger people. This means you may need to stop collecting responses from older people sooner than you will for other age groups. 

Step 4: Adjust your quotas if needed

As previously mentioned, it’s common for some quotas to fill up more quickly than others. This means you may have to actively seek out respondents who fit the profile for unmet quotas. (Some audience groups are particularly difficult to recruit.) 

For example, while you may know that the distribution of people who use an adult product brand is 70 percent males and 30 percent females, it may be difficult to find 300 females (based on a sample size of 1,000) who are willing to participate in the study. In such a scenario, you may want to adjust your quota to allow more male users to participate in the study. 

Once you’ve completed the survey, you may then use other methods, such as weighting — a correction technique that allows you to add more power to the responses of the groups with unmet quotas — to adjust the results.

Advantages of quota sampling

There are numerous reasons why researchers may prefer to use quota sampling:

• Accuracy: With quota sampling, all proportions are accounted for and represented in the sample. This helps prevent over-representation or under-representation of a specific group in the general population.
• Convenience: Compared to other sampling methods, quota sampling is fairly straightforward, and the results are easy to understand and analyze. Once quotas are set, researchers can easily get data on a given population until all quotas are met. 
• Affordability: The cost of conducting quota sampling is generally lower compared to other methods. For example, researchers can collect data completely through online forms, and there’s usually no traveling required.

Disadvantages of quota sampling

Although quota sampling has its pros, it may still present some challenges:

• Since only specific population traits are considered when samples are stratified into subgroups, all characteristics may get equal representation. Quota sampling may miss nuances in the wider population. 
• It may create space for bias during selection. Since quota sampling depends on the researcher’s judgment, the researcher may select based on convenience, thereby introducing bias.
• Dividing the entire population into mutually exclusive groups may be difficult because people can belong to two or more groups. Hence, it affects the data-gathering process when people don’t fit in one distinct group.

Faster quota sampling with Jotform

Survey tools like Jotform make it easier to conduct quota sampling. Jotform allows you to create simple or complex online surveys. You can quickly view the results as graphs, percentages, or charts in real time. Once quotas start to fill up, you can also implement conditional logic to stop receiving responses from that audience group. Get started with Jotform today. 

Photo by: drobotdean