A population is the entire collection of elements, parts, people, etc. in the study of interest. It is rare that you have the entire population available for your study. So, in order to collect data, you must take a sample.

A sample is a subset from the population of interest on which you will make observations once you have collected it.

Not only must you decide what sampling method to use, you must first clearly define the population, including the timeframe when you will be collecting your sample.

But what sampling method should you use?

There are two main costs that must be addressed in your choice: the process (and therefore the cost of collecting your sample) and the costs associated with making an incorrect decision due to systematic bias based on your sampling method.

In this short video, Nicola Petty, an expert in teaching and learning statistics, discusses the five main methods for taking a sample and their advantages and disadvantages.

She addresses the ease of taking a sample and the propensity for systematic bias for each sampling method:

  • Simple Random
  • Convenience
  • Systematic
  • Cluster
  • Stratified

You can watch Dr. Petty’s video here!