CRIM 220 Lecture Notes - Lecture 8: Random Number Generation, Sample Size Determination, Statistical Parameter
Document Summary
Often not possible to study entire population. Often not necessary to study entire population. Can specify for each element of population probability that it will be included in sample. No way to estimate probability each element has of being included in sample. Each member of population has known & equal chance of being selected. Contains same kind of variations existing in population. To select set of elements from population in such a way that descriptions of those elements accurate portray parameters of total population from which elements selected. Each elements of population has equal chance of selection. Sampling frame - list of 30,000 student #s. Sampling frame - identifies fact that population may not be 100% representative. Closest we can get to population list. Average of infinite # of sample statistics = E. g. average approval rating for infinite # of samples of 100 students = actual approval rating for population of 30k students.