The process of statistical sampling involves selecting a collection of individuals from a population. The way that we do this selection is very important. The manner in which we select our sample determines the type of sample that we have. Among the wide variety of types of statistical samples, the easiest type of sample to form is called a convenience sample.

## Definition of Convenience Samples

A convenience sample is formed when we select elements from a population on the basis of what elements are easy to obtain. Sometimes a convenience sample is called a grab sample as we essentially grab members from the population for our sample. This is a type of sampling technique that does not rely upon a random process, such as we see in a simple random sample, to generate a sample.

## Examples of Convenience Samples

To illustrate the idea of a convenience sample, we will think of several examples. It is really not very hard to do this. Just think of the easiest way to find representatives for a particular population. There is a high likelihood that we have formed a convenience sample.

- To determine the proportion of green M&Ms produced by a factory, we count the number of green M&Ms in our hands that we took out of the package.
- To find the mean height of all third-grade students in a school district, we measure the first five students who are dropped off in the morning by their parents.
- In order to know the mean value of homes in our town, we average the value of our home with our neighbors' homes.
- Someone wants to determine which candidate is likely to win an upcoming election, and so she asks everyone in her circle of friends who they intend to vote for.
- A student is working on a survey of students' attitudes toward college administrators, and so he talks to his roommate and other people on the floor of his residence hall.

## Problems with Convenience Samples

As indicated by their name, convenience samples are definitely easy to obtain. There is virtually no difficulty in selecting members of the population for a convenience sample. However, there is a price to pay for this lack of effort: convenience samples are virtually worthless in statistics.

The reason that a convenience sample cannot be used for applications in statistics is that we are not assured that it is representative of the population that it was selected from. If all of our friends share the same political leanings, then asking them who they intend to vote for in an election tells us nothing about how people across the country would vote.

Furthermore, if we think about the reason for random sampling, we should see another reason why convenience samples are not as good as other sampling designs. Since we do not have a random procedure to select the individuals in our sample, although our sample is likely to be biased. A randomly selected sample will do a better job of limiting bias.