ABSTRACT
A sample consisting of people encountered on a downtown street during the noon hour would almost certainly contain a higher proportion of employed women than in the population. Often, people try to eliminate systematic bias by carefully selecting individuals who seem to be typical of the population. The idea of simple random sampling is that every individual in the population has the same chance of being selected for the sample. Randomization does not guarantee, of course, that the sample will look like the population. Stratified sampling is a way of ensuring that a random sample will represent various groups in the same proportions as in the whole population Variables used in studies of people, for example, include height, hair color, attitude toward nuclear power as indicated by response to a question on a particular survey, age at death, and so on.
