Texas A&M University (TAMU) STAT 201 Elementary Statistical Inference Exam 1 Practice

Stratified sampling keeps representation by splitting the population into subgroups that are similar within themselves but different from each other, and then drawing a random sample from each subgroup. This approach ensures that every subgroup contributes to the overall sample, which can reduce sampling error when outcomes vary by subgroup.

The statement that describes dividing the population into subgroups and taking a random sample from each subgroup matches this idea exactly. If you were to divide into subgroups and include all members from some of them, that would describe sampling entire clusters, which is a different design. Convenience sampling and systematic sampling are other methods that don’t fit stratified sampling: convenience samples are drawn from who’s easiest to reach, and systematic samples pick every kth individual.