This tutorial will explain quota sampling by exploring the following topics:
Like a stratified random sampling, the sampling frame (all the people or subjects who could be sampled) is broken down into smaller subgroups. These subgroups could be gender, race, graduating class, etc. The idea is to see the differences between the subgroups. However, unlike stratified random sampling, quota sampling deals with selecting a predetermined number from each group. The difference is that it doesn't need to be random.
Typically, the purpose of quota sampling is to ensure that groups in the sample are roughly proportionally represented in the sample. Much as they are in the population.
Suppose that a high school adopted new healthy lunch options and they want to solicit student feedback on them. The school has 100 freshman, 110 sophomores, 120 juniors, and 90 seniors.
Their goal is to sample 10% of the school to obtain feedback on that. How would you do that?One way to do it would be to just sample 10% of each class so, 10 freshman, 11 sophomores, 12 juniors, and 9 seniors. It's not important how these students are obtained, as long as the number is reached.
The numbers 10, 11, and 12 are the “quota”.
Suppose the sample was taken by reaching out to the first 10 freshman, 11 sophomores, 12 juniors, and 9 senior that entered the lunchroom. How might this affect the feedback?
The percent of students that like the lunch options might be over-represented because they happened to be the first ones to lunch. They might simply like the lunch options more than the average student.
The results might over represent systematically the percent of students that enjoy these options and the sample would be biased because they weren’t randomly selected. This would be a type of convenience sample at lunchtime where the students were easy to obtain.
The benefits of doing a quota sample are:
Easy to evaluate if sophomores prefer the lunch options more than seniors do.
Since the selection process isn't random, the sample may not be representative of the population, therefore making it difficult to generalize the results of the study to the population. It cannot be emphasized enough how big of a problem this is.
The goal is to be able to generalize the results of a study to the population at large. If the selection process isn't random, there's no real guarantee that this can be done.
It is strongly suggested that, if possible, stratified random sample is used as opposed to a quota sample.
Quota sampling, as in stratified random sampling, the population is broken down into strata. Then, a sample is taken from within each of those strata. The major difference is that the samples within the strata don't need to be randomly selected.
Benefits of quota sampling are that it is easy to conduct and it’s easy to see results according to categories. The biggest drawback of quota sampling is the limited ability to generalize results to the population.
Source: This work is adapted from Sophia author jonathan Osters.
A sampling method where a certain, predetermined number of individuals are taken from each of several different classes of the population. The selection method does not need to be random, which may not result in a representative sample of the population.