ABSTRACT
This chapter begins with a discussion of some fundamental concepts in sampling. Sampling in public administration research is used because it is more efficient than studying a full population. Sampling can lower the cost and improve the efficiency of measurement activities. The five fundamental considerations of the sampling process result in many different possible combinations in sample design. Sample design is based on these choices: probability or non- probability, single-unit or unit clusters, stratified or unstratified, equal-unit probability or weighted probability, and single-stage or multistage sampling. Decision makers must take care to avoid introducing bias and sampling or nonsampling error into the sample design and selection process. According to the central limit theorem, as the size of the sample increases, the distribution of the sample measurements tends to take on the shape of normal distribution. The chapter concludes with a review of several statistical procedures relating to sampling.
