What is stratified sampling?

• A. Dividing population into homogeneous subgroups and sampling from each
• B. Selecting individuals at random without regard to subgroup
• C. Sampling only from the largest subgroup
• D. Using only one simple random sample from the entire population

Answer: (A)

Stratified sampling means dividing the population into distinct subgroups that are homogeneous within, and then drawing samples from each subgroup. The idea is that people within a subgroup are similar on the characteristic that matters for the study, so sampling from every subgroup ensures the overall sample reflects the population’s structure and reduces sampling error compared with taking a single random sample from the whole group.

This approach is particularly helpful when subgroups differ in the outcome of interest—for example, if you’re studying average income and there are clearly different income ranges across regions or age groups, stratifying by region or age helps you capture those differences accurately. After defining the strata (which must be mutually exclusive and cover everyone), you sample from each stratum, either in proportion to its size or with a fixed number from each group, depending on the study’s goals. It can be more precise but may be more costly than simple random sampling.