Prestratification

Prestratification is a statistical sampling technique used to improve the efficiency and representativeness of a sample drawn from a population. It involves dividing the population into distinct subgroups, known as strata, before the sample is selected. These strata are mutually exclusive and collectively exhaustive, meaning every member of the population belongs to exactly one stratum. The division is typically based on one or more characteristics that are believed to be relevant to the variable of interest.

The primary goal of prestratification is to ensure that specific subgroups within the population are adequately represented in the sample, even if they are small. By oversampling or undersampling certain strata, researchers can achieve a more precise estimate of population parameters for the entire population, or to make more accurate comparisons between strata. For example, in a survey about educational attainment, a population might be stratified by geographic region, socioeconomic status, or type of school attended.

Once the strata are defined, sampling is conducted independently within each stratum. The sampling method within each stratum can vary, but it is often proportional to the size of the stratum in the population. Alternatively, disproportionate sampling can be used to give more weight to smaller or more important strata. The results from each stratum are then combined, usually through weighted averaging, to produce an estimate for the overall population. This process helps to reduce sampling error and provides a more reliable representation of the population's characteristics.