Sampling Methods

There are four basic methods of acquiring a sample.

Random Sampling
Ideally, all members of a population have an equal chance of being used as a member of the sample. We like to pretend this is true even though we don't use the same person twice so each subsequent chosen individual had a better chance of being picked than the previous individual(s).


 * For example, if the population was all people who play the lottery, then using the lottery winners as the sample would be a form of random sampling because each individual had an equal chance of winning.

Stratified Sampling
The population is divided into strata (or groups) based on some characteristic(s). Then individuals are selected from each strata (or group) proportionate to those in the population. This helps make your sample more representative of the target population.


 * For example, if we wanted to do a national survey, we might divide people into racial categories. Then based on the 2010 Census, we know that our sample should be 72.4% white, 12.6% black/African-American, 0.9% American Indian/Alaska Native, 4.8% Asian, 0.2% Native Hawaiian/Pacific Islander, 6.2% other, and 2.9% two or more races.

Purposive Sampling
Sampling with a purpose in mind. This is performed when you have at least one predefined group of interest. Only individuals who appear to fit in that group are used in the sample.


 * For example, you may want to study Asian males between the ages of 20 and 30 so you only seek people who appear to fit that profile.

Incidental, Chunk, Accidental, or Convenience Sampling
This is the weakest sampling method. The method involves using individuals who are easy to access as your sample.


 * For example, you may want to know the political views of all US Citizens. Instead of trying to ask people from all over the country, you may only ask students at your college/university because it's easier.