Random Sampling vs. Quota Sampling

Sampling Populations

If I were to tell you that 85% of Americans love chocolate cake, you may very well agree with that because, I mean, who doesn't love chocolate cake? While a statistic like that may be true, it would be impossible for me to ask every single American their opinion on chocolate cake. In fact, that would require me to ask over 300 million people.

Instead, when we see statistics like the one just mentioned, the attitudes or opinions they reflect are often drawn from a sample, or selected portion of the greater whole. There are many ways in which a sample can be taken. In this lesson, we will discuss two ways in which samples are taken. These ways are random sampling and quota sampling.

Random Sampling

The most accurate way to obtain information for a large group of subjects is by using probability sampling, where samples are selected in such a way as to be representative of a population. When we say representative, we mean all the types of people we could expect to see living in that population. For instance, if we were surveying video game playing behavior, people over 65 may not be included because that age and above isn't typically representative of people who like to play video games. Probability sampling provides the most valid, or credible, results because it reflects the characteristics of the population from which they are selected.

A common type of probability sampling is known as random sampling. The term 'random' has a very precise meaning. In random sampling, each individual in the population of interest has an equal likelihood of selection. This means that you can't just collect responses on the street and have a random sample because not everyone that can be selected may be on that street, or that street may be in an area that only houses people with certain characteristics. For a truly random sample, think of your whole population being placed into a bag and you closing your eyes and pulling your participants out to participate in your research. The key to random selection is that there is no bias involved in the selection of the sample. Any variation between the sample characteristics and the population characteristics is only a matter of chance.

It's important to understand that the assumption of an equal chance of selection means that sources such as a telephone book or voter registration lists are not adequate for providing a random sample of a community. In both these cases, there will be a number of residents whose names are not listed. In fact, lists of registered drivers is now commonly used because those lists tend to be more representative of a population. Telephone surveys get around this problem by random-digit dialing, but that still assumes that everyone in the population has a telephone.

One of the most classic historical examples of a random sample bias was in the 1936 presidential election where Literary Digest predicted that Alfred Landon would beat future president Franklin Roosevelt. The survey sample suffered from undercoverage of low-income voters, who tended to be Democrats and were thus voting for Roosevelt. You might be asking yourself, how did this happen? Well, the survey relied on a convenience sample, drawn from telephone directories and car registration lists. In 1936, people who owned cars and telephones tended to be more affluent and not the low-income Democrats who were voting for Roosevelt.

Quota Sampling

Sometimes, poll or survey results are obtained using non-probability sampling methods, where those selected are not truly representative of a given population. In general, non-probability samples are less desirable than probability samples. However, a researcher may not be able to obtain a random sample because it may be too expensive, or a researcher may not care about generalizing to a larger population. Nevertheless, the validity of non-probability samples can be increased by trying to approximate random selection and by eliminating as many sources of bias as possible. One method that is used to approximate random selection is by quota sampling.

In a quota sample, a researcher deliberately sets the proportions of levels of members chosen within the sample. This is generally done to ensure the inclusion of a particular segment of the population that may otherwise be overlooked. The researcher sets a quota, independent of population characteristics, so the proportions may differ from the actual proportion in the population.