In stratified random sampling, how are samples drawn?

• A. Dividing the population into subpopulations and sampling only from a single subpopulation
• B. Convenience sampling within each subpopulation
• C. Selecting entire subpopulations without randomization
• D. Dividing the population into strata and random samples from each stratum

Answer: (D)

Stratified random sampling works by dividing the population into strata—groups that share a similar characteristic and are more homogeneous within than between groups—and then drawing random samples from each stratum. This ensures that every subgroup is represented in the overall sample and often reduces sampling error because variation within strata is smaller than variation across the entire population. You would sample from each stratum using a random method, and you can choose to allocate samples proportionally to each stratum’s size or use an equal number from each stratum, depending on the design.

Options that sample only from one subpopulation, use convenience sampling within subpopulations, or select entire subpopulations without randomization don’t meet the principle of stratified random sampling, which combines division into strata with random sampling from each.