At the heart of the practice of statistics is statistical sampling. In this process we select a subgroup called a sample from a population and study this subgroup. The goal is to use the sample to say something meaningful about the population. Not all samples are the same. One type of sample is called a stratified sample.

### What Is a Stratified Sample?

Sometimes a population naturally splits into several non-overlapping portions.

We envision these pieces of the population as strata – or layers. A stratified sample recognizes these layers of our population. Our sample has some restrictions imposed upon it. We make sure that our statistical sample contains representatives from each of the strata.

### Reasons for Using a Stratified Sample

A simple random sample is typically the goal in any statistical study. However, our study may intend to compare or contrast members from each stratum of the population. In theory a simple random sample could misrepresent the composition of the population. Use of a stratified sample will ensure that representatives from every portion of the population will be selected.

### Example of a Stratified Sample

It helps to think through a few examples of what constitutes a stratified sample. One example of this type of sample would be to consider a population that consists of a group of people. We could stratify this population on the basis of gender.

Males would constitute one stratum and females a second. A simple random sample from each of these strata would then form a stratified sample.

Another example of a stratified sample would start with a population of high school students. These students could be stratified by their academic standing: freshman, sophomore, junior or senior.