ABSTRACT
What is the average age of single mothers in the United States? How often during the day does a hyperactive child leave his or her desk? What problems are faced by persons who have been sexually assaulted? All of these are questions about universes of people, time periods, or behaviors. The obvious way of answering them would be to observe the population, or entire universe of elements being examined. But such an approach is clearly impractical and, in many instances, impossible. To answer these questions using the entire population, we would have to interview all single mothers, observe the hyperactive child continuously throughout the day, and assess all persons who have been sexually assaulted. Instead, we must be satisfied with observing a subset of the populations in which we have an interest. Based on our observations of the subset, or sample, we want to make inferences about the characteristics of the entire population from which that sample was taken.
