When carrying out a stratified sampling, the researchers split the population into different groups. The resulting sample is a consequence of some members from each group who are chosen randomly. This method ensures representation for every individual in each group in the sample, and it avoids a biased result.
