sampling, stratified

n. Stratified sampling is a method of selecting members for a sample so that the relevant characteristics of the sample are similar to those of the larger target population. The population is divided into nonoverlapping and collectively exhaustive subpopulations or strata (the number of strata depends on the target of the study and the relevant grouping characteristics). Sample members are then randomly selected from each stratum or subgroup in proportion to the population. The members of each stratum are pooled to form the overall sample. For example, if you were interested in studying political attitudes on a specific college campus, you would want to ensure that your study sampled a representative number of freshmen, sophomores, juniors, and seniors. So, you would divide your population (all students) into strata (class levels) and randomly select your sample using representative numbers from each subgroup. Stratified sampling is necessary for making valid inferences and generalizing conclusions to the target population. – BJM