The population of all the data is represented by a simple random sample. Using a stratified random sample, the population is divided into smaller groups, or strata, according to shared traits.

Every technique has it's own advantages and disadvantages to know it better you should learn about them properly. Learn it from Data Analytics Training.

Contrary to basic random samples, stratified random samples are applied to populations that are easily divided into various divisions or subsets. These groups are based on specific criteria, and components from each are then randomly selected in proportion to the size of the group compared to the population. S&P 500 firms in the aforementioned example may have been split down by industry or headquarter area.

The magnitude of the selections from each group will depend on its proportion to the total population under this sampling procedure. The strata must not overlap, according to the researchers. Every point in the population must be mutually exclusive and can only belong to one stratum. It is more likely that some data will be included when strata overlap, which will skew the sample.