In many scientific inquiries, it is impossible to individually appraise all the elements that comprise a population of interest. Instead, there is a need to infer collective properties of the whole population from a select subset of its elements. Probability sampling is an approach to selecting elements from a fixed population in such a way that
- elements are selected by a random process,
- every element has a nonzero chance of selection, and
- the relative frequency with which an element is included in a sample is deducible.
A collection of elements drawn in such a way is referred to as a probability sample. The first condition imparts a degree of objectivity to a probability sample and, in combination with the other two conditions, secures a basis for statistical inferences concerning ...