Stratified sampling
The act of dividing a larger population
into subgroups, using systematic
sampling, then taking a random
sample from each subgroup. Random sampling can frequently minimize the
sampling error in the population. This in turn
increases the precision of any estimation
methods used.
You MUST be sure to select relevant
variables for the grouping process. The information, or data, must be as
accurate as possible.
This is a method of sampling with any given
population could vary significantly depending on the grouping and how
uniform the grouping is.
Also see Sampling
Bias.
