This chapter deals with estimation of population quantities in surveys with a known sampling design, specified (controlled) by the designer of the survey. Sampling theory treats the values of variables as constants, and chance decides whether or not they are observed for any given member of the population. By sampling design, we set the joint probabilities associated with these observations. In Sections 3.7 and 3.7.1, pursuing efficient estimation of population quantities, we bring into play regression and related models, making use of the experience with them in the infinite population setting of Chapters 1 and 2.
KeywordsSampling Design Sampling Variance Simple Random Sampling Ordinary Regression Auxiliary Information
Unable to display preview. Download preview PDF.